Speech Recognition

Languages 99

Generations 3 0 External Speechmatics model. Language-agnostic adapter deployed per tenant. 1 Dedicated GPU per language per tenant. No shared infrastructure or dynamic routing. 2 AED on Triton via gRPC request/response. Shared across tenants with dynamic routing. 3 TDT on Triton via gRPC request/response. Shared across tenants with dynamic routing.

Tenants 14

Production Models

eu/aleris prod-weu/aleris Routing config 46 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da prd-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de prd-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en prd-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es prd-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr prd-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual prd-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de prd-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en prd-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch prd-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no prd-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv prd-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
speechmatics-adapter eu/aleris/speechmatics-adapter	0 External Speechmatics model. Language-agnostic adapter deployed per tenant.	Routing enabled via `ml-service-routing`	ar → Ambient bg → Ambient cmn → Ambient cs → Ambient el → Ambient et → Ambient fi → Ambient he → Ambient hi → Ambient hr → Ambient hu → Ambient it → Ambient ja → Ambient ko → Ambient lt → Ambient lv → Ambient mr → Ambient ms → Ambient mt → Ambient nl → Ambient no → Ambient pl → Ambient pt → Ambient ro → Ambient ru → Ambient sk → Ambient sl → Ambient sv → Ambient sw → Ambient ta → Ambient th → Ambient tr → Ambient uk → Ambient ur → Ambient vi → Ambient yue → Ambient	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input.

eu/base prod-weu/base Routing config 55 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da prd-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline da-x-beta → Ambient da-x-beta → Dictation da-x-beta → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de prd-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline de-CH-x-beta → Ambient de-CH-x-beta → Dictation de-CH-x-beta → Offline de-x-beta → Ambient de-x-beta → Dictation de-x-beta → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en prd-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-AU-x-beta → Ambient en-AU-x-beta → Dictation en-AU-x-beta → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-GB-x-beta → Ambient en-GB-x-beta → Dictation en-GB-x-beta → Offline en-US-x-beta → Ambient en-US-x-beta → Dictation en-US-x-beta → Offline en-x-beta → Ambient en-x-beta → Dictation en-x-beta → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es prd-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr prd-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline fr-CH-x-beta → Ambient fr-CH-x-beta → Dictation fr-CH-x-beta → Offline fr-x-beta → Ambient fr-x-beta → Dictation fr-x-beta → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual prd-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de prd-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en prd-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch prd-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no prd-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv prd-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
speechmatics-adapter eu/base/speechmatics-adapter	0 External Speechmatics model. Language-agnostic adapter deployed per tenant.	Routing enabled via `ml-service-routing`	ar → Ambient bg → Ambient cmn → Ambient cs → Ambient el → Ambient et → Ambient fi → Ambient he → Ambient hi → Ambient hr → Ambient hu → Ambient it → Ambient ja → Ambient ko → Ambient lt → Ambient lv → Ambient mr → Ambient ms → Ambient mt → Ambient nl → Ambient no → Ambient pl → Ambient pt → Ambient ro → Ambient ru → Ambient sk → Ambient sl → Ambient sv → Ambient sw → Ambient ta → Ambient th → Ambient tr → Ambient uk → Ambient ur → Ambient vi → Ambient yue → Ambient	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input.

eu/bighand prod-weu/bighand Routing config 10 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da prd-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de prd-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en prd-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es prd-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr prd-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual prd-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de prd-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en prd-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch prd-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no prd-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv prd-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.

eu/bupa prod-weu/bupa Routing config 46 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da prd-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de prd-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en prd-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es prd-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr prd-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual prd-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de prd-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en prd-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch prd-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no prd-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv prd-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
speechmatics-adapter eu/bupa/speechmatics-adapter	0 External Speechmatics model. Language-agnostic adapter deployed per tenant.	Routing enabled via `ml-service-routing`	ar → Ambient bg → Ambient cmn → Ambient cs → Ambient el → Ambient et → Ambient fi → Ambient he → Ambient hi → Ambient hr → Ambient hu → Ambient it → Ambient ja → Ambient ko → Ambient lt → Ambient lv → Ambient mr → Ambient ms → Ambient mt → Ambient nl → Ambient no → Ambient pl → Ambient pt → Ambient ro → Ambient ru → Ambient sk → Ambient sl → Ambient sv → Ambient sw → Ambient ta → Ambient th → Ambient tr → Ambient uk → Ambient ur → Ambient vi → Ambient yue → Ambient	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input.

eu/celo prod-weu/celo Routing config 10 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da prd-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de prd-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en prd-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es prd-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr prd-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual prd-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de prd-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en prd-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch prd-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no prd-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv prd-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.

eu/dedalus prod-weu/dedalus Routing config 46 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da prd-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de prd-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en prd-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es prd-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr prd-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual prd-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de prd-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en prd-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch prd-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no prd-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv prd-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
speechmatics-adapter eu/dedalus/speechmatics-adapter	0 External Speechmatics model. Language-agnostic adapter deployed per tenant.	Routing enabled via `ml-service-routing`	ar → Ambient bg → Ambient cmn → Ambient cs → Ambient el → Ambient et → Ambient fi → Ambient he → Ambient hi → Ambient hr → Ambient hu → Ambient it → Ambient ja → Ambient ko → Ambient lt → Ambient lv → Ambient mr → Ambient ms → Ambient mt → Ambient nl → Ambient no → Ambient pl → Ambient pt → Ambient ro → Ambient ru → Ambient sk → Ambient sl → Ambient sv → Ambient sw → Ambient ta → Ambient th → Ambient tr → Ambient uk → Ambient ur → Ambient vi → Ambient yue → Ambient	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input.

eu/medatixx prod-weu/medatixx Routing config 10 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da prd-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de prd-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en prd-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es prd-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr prd-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual prd-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de prd-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en prd-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch prd-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no prd-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv prd-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.

eu/trifork prod-weu/trifork Routing config 46 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da prd-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de prd-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en prd-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es prd-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr prd-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual prd-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de prd-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en prd-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch prd-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no prd-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv prd-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
speechmatics-adapter eu/trifork/speechmatics-adapter	0 External Speechmatics model. Language-agnostic adapter deployed per tenant.	Routing enabled via `ml-service-routing`	ar → Ambient bg → Ambient cmn → Ambient cs → Ambient el → Ambient et → Ambient fi → Ambient he → Ambient hi → Ambient hr → Ambient hu → Ambient it → Ambient ja → Ambient ko → Ambient lt → Ambient lv → Ambient mr → Ambient ms → Ambient mt → Ambient nl → Ambient no → Ambient pl → Ambient pt → Ambient ro → Ambient ru → Ambient sk → Ambient sl → Ambient sv → Ambient sw → Ambient ta → Ambient th → Ambient tr → Ambient uk → Ambient ur → Ambient vi → Ambient yue → Ambient	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input.

fr/fr prd-par01/fr Routing config 46 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da prd-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de prd-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en prd-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es prd-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr prd-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual prd-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de prd-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en prd-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch prd-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no prd-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv prd-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
speechmatics-adapter fr/fr/speechmatics-adapter	0 External Speechmatics model. Language-agnostic adapter deployed per tenant.	Routing enabled via `ml-service-routing`	ar → Ambient bg → Ambient cmn → Ambient cs → Ambient el → Ambient et → Ambient fi → Ambient he → Ambient hi → Ambient hr → Ambient hu → Ambient it → Ambient ja → Ambient ko → Ambient lt → Ambient lv → Ambient mr → Ambient ms → Ambient mt → Ambient nl → Ambient no → Ambient pl → Ambient pt → Ambient ro → Ambient ru → Ambient sk → Ambient sl → Ambient sv → Ambient sw → Ambient ta → Ambient th → Ambient tr → Ambient uk → Ambient ur → Ambient vi → Ambient yue → Ambient	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input.

us/base prod-us/base Routing config 55 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da prd-ric02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline da-x-beta → Ambient da-x-beta → Dictation da-x-beta → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de prd-ric02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline de-CH-x-beta → Ambient de-CH-x-beta → Dictation de-CH-x-beta → Offline de-x-beta → Ambient de-x-beta → Dictation de-x-beta → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en prd-ric02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-AU-x-beta → Ambient en-AU-x-beta → Dictation en-AU-x-beta → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-GB-x-beta → Ambient en-GB-x-beta → Dictation en-GB-x-beta → Offline en-US-x-beta → Ambient en-US-x-beta → Dictation en-US-x-beta → Offline en-x-beta → Ambient en-x-beta → Dictation en-x-beta → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es prd-ric02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr prd-ric02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline fr-CH-x-beta → Ambient fr-CH-x-beta → Dictation fr-CH-x-beta → Offline fr-x-beta → Ambient fr-x-beta → Dictation fr-x-beta → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual prd-ric02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch prd-ric02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no prd-ric02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv prd-ric02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
speechmatics-adapter us/base/speechmatics-adapter	0 External Speechmatics model. Language-agnostic adapter deployed per tenant.	Routing enabled via `ml-service-routing`	ar → Ambient bg → Ambient cmn → Ambient cs → Ambient el → Ambient et → Ambient fi → Ambient he → Ambient hi → Ambient hr → Ambient hu → Ambient it → Ambient ja → Ambient ko → Ambient lt → Ambient lv → Ambient mr → Ambient ms → Ambient mt → Ambient nl → Ambient no → Ambient pl → Ambient pt → Ambient ro → Ambient ru → Ambient sk → Ambient sl → Ambient sv → Ambient sw → Ambient ta → Ambient th → Ambient tr → Ambient uk → Ambient ur → Ambient vi → Ambient yue → Ambient	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input.

staging-eu/base staging-eu/base Routing config 99 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da stg-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline da-x-beta → Ambient da-x-beta → Dictation da-x-beta → Offline da-x-noformat → Ambient da-x-noformat → Dictation da-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de stg-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline de-CH-x-beta → Ambient de-CH-x-beta → Dictation de-CH-x-beta → Offline de-CH-x-noformat → Ambient de-CH-x-noformat → Dictation de-CH-x-noformat → Offline de-x-beta → Ambient de-x-beta → Dictation de-x-beta → Offline de-x-noformat → Ambient de-x-noformat → Dictation de-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en stg-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-AU-x-beta → Ambient en-AU-x-beta → Dictation en-AU-x-beta → Offline en-AU-x-noformat → Ambient en-AU-x-noformat → Dictation en-AU-x-noformat → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-GB-x-beta → Ambient en-GB-x-beta → Dictation en-GB-x-beta → Offline en-GB-x-noformat → Ambient en-GB-x-noformat → Dictation en-GB-x-noformat → Offline en-US-x-beta → Ambient en-US-x-beta → Dictation en-US-x-beta → Offline en-x-beta → Ambient en-x-beta → Dictation en-x-beta → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline en-x-noformat → Ambient en-x-noformat → Dictation en-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es stg-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline es-x-noformat → Ambient es-x-noformat → Dictation es-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr stg-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline fr-CH-x-beta → Ambient fr-CH-x-beta → Dictation fr-CH-x-beta → Offline fr-CH-x-noformat → Ambient fr-CH-x-noformat → Dictation fr-CH-x-noformat → Offline fr-x-beta → Ambient fr-x-beta → Dictation fr-x-beta → Offline fr-x-noformat → Ambient fr-x-noformat → Dictation fr-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual stg-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de stg-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en stg-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch stg-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-mock stg-par02/mlservices/asr-aed-mock	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no stg-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv stg-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
speechmatics-adapter staging-eu/base/speechmatics-adapter	0 External Speechmatics model. Language-agnostic adapter deployed per tenant.	Routing enabled via `ml-service-routing`	ar → Ambient ar-x-noformat → Ambient bg → Ambient bg-x-noformat → Ambient cmn → Ambient cmn-x-noformat → Ambient cs → Ambient cs-x-noformat → Ambient el → Ambient el-x-noformat → Ambient et → Ambient et-x-noformat → Ambient fi → Ambient fi-x-noformat → Ambient he → Ambient he-x-noformat → Ambient hi → Ambient hi-x-noformat → Ambient hr → Ambient hr-x-noformat → Ambient hu → Ambient hu-x-noformat → Ambient it → Ambient it-x-noformat → Ambient ja → Ambient ja-x-noformat → Ambient ko → Ambient ko-x-noformat → Ambient lt → Ambient lv → Ambient lv-x-noformat → Ambient mr → Ambient mr-x-noformat → Ambient ms → Ambient ms-x-noformat → Ambient mt → Ambient mt-x-noformat → Ambient nl → Ambient nl-x-noformat → Ambient no → Ambient no-x-noformat → Ambient pl → Ambient pl-x-noformat → Ambient pt → Ambient pt-x-noformat → Ambient ro → Ambient ro-x-noformat → Ambient ru → Ambient ru-x-noformat → Ambient sk → Ambient sk-x-noformat → Ambient sl → Ambient sl-x-noformat → Ambient sv → Ambient sv-x-noformat → Ambient sw → Ambient sw-x-noformat → Ambient ta → Ambient ta-x-noformat → Ambient th → Ambient th-x-noformat → Ambient tr → Ambient tr-x-noformat → Ambient uk → Ambient uk-x-noformat → Ambient ur → Ambient ur-x-noformat → Ambient vi → Ambient vi-x-noformat → Ambient yue → Ambient yue-x-noformat → Ambient	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input.

stg/base stg-par01/base Routing config 99 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da stg-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline da-x-beta → Ambient da-x-beta → Dictation da-x-beta → Offline da-x-noformat → Ambient da-x-noformat → Dictation da-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de stg-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline de-CH-x-beta → Ambient de-CH-x-beta → Dictation de-CH-x-beta → Offline de-CH-x-noformat → Ambient de-CH-x-noformat → Dictation de-CH-x-noformat → Offline de-x-beta → Ambient de-x-beta → Dictation de-x-beta → Offline de-x-noformat → Ambient de-x-noformat → Dictation de-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en stg-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-AU-x-beta → Ambient en-AU-x-beta → Dictation en-AU-x-beta → Offline en-AU-x-noformat → Ambient en-AU-x-noformat → Dictation en-AU-x-noformat → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-GB-x-beta → Ambient en-GB-x-beta → Dictation en-GB-x-beta → Offline en-GB-x-noformat → Ambient en-GB-x-noformat → Dictation en-GB-x-noformat → Offline en-US-x-beta → Ambient en-US-x-beta → Dictation en-US-x-beta → Offline en-x-beta → Ambient en-x-beta → Dictation en-x-beta → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline en-x-noformat → Ambient en-x-noformat → Dictation en-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es stg-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline es-x-noformat → Ambient es-x-noformat → Dictation es-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr stg-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline fr-CH-x-beta → Ambient fr-CH-x-beta → Dictation fr-CH-x-beta → Offline fr-CH-x-noformat → Ambient fr-CH-x-noformat → Dictation fr-CH-x-noformat → Offline fr-x-beta → Ambient fr-x-beta → Dictation fr-x-beta → Offline fr-x-noformat → Ambient fr-x-noformat → Dictation fr-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual stg-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de stg-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en stg-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch stg-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-mock stg-par02/mlservices/asr-aed-mock	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no stg-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv stg-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
speechmatics-adapter stg/base/speechmatics-adapter	0 External Speechmatics model. Language-agnostic adapter deployed per tenant.	Routing enabled via `ml-service-routing`	ar → Ambient ar-x-noformat → Ambient bg → Ambient bg-x-noformat → Ambient cmn → Ambient cmn-x-noformat → Ambient cs → Ambient cs-x-noformat → Ambient el → Ambient el-x-noformat → Ambient et → Ambient et-x-noformat → Ambient fi → Ambient fi-x-noformat → Ambient he → Ambient he-x-noformat → Ambient hi → Ambient hi-x-noformat → Ambient hr → Ambient hr-x-noformat → Ambient hu → Ambient hu-x-noformat → Ambient it → Ambient it-x-noformat → Ambient ja → Ambient ja-x-noformat → Ambient ko → Ambient ko-x-noformat → Ambient lt → Ambient lv → Ambient lv-x-noformat → Ambient mr → Ambient mr-x-noformat → Ambient ms → Ambient ms-x-noformat → Ambient mt → Ambient mt-x-noformat → Ambient nl → Ambient nl-x-noformat → Ambient no → Ambient no-x-noformat → Ambient pl → Ambient pl-x-noformat → Ambient pt → Ambient pt-x-noformat → Ambient ro → Ambient ro-x-noformat → Ambient ru → Ambient ru-x-noformat → Ambient sk → Ambient sk-x-noformat → Ambient sl → Ambient sl-x-noformat → Ambient sv → Ambient sv-x-noformat → Ambient sw → Ambient sw-x-noformat → Ambient ta → Ambient ta-x-noformat → Ambient th → Ambient th-x-noformat → Ambient tr → Ambient tr-x-noformat → Ambient uk → Ambient uk-x-noformat → Ambient ur → Ambient ur-x-noformat → Ambient vi → Ambient vi-x-noformat → Ambient yue → Ambient yue-x-noformat → Ambient	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input.

dev-weu/base dev-weu/base Routing config 99 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da dev-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline da-x-beta → Ambient da-x-beta → Dictation da-x-beta → Offline da-x-noformat → Ambient da-x-noformat → Dictation da-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de dev-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline de-CH-x-beta → Ambient de-CH-x-beta → Dictation de-CH-x-beta → Offline de-CH-x-noformat → Ambient de-CH-x-noformat → Dictation de-CH-x-noformat → Offline de-x-beta → Ambient de-x-beta → Dictation de-x-beta → Offline de-x-noformat → Ambient de-x-noformat → Dictation de-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en dev-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-AU-x-beta → Ambient en-AU-x-beta → Dictation en-AU-x-beta → Offline en-AU-x-noformat → Ambient en-AU-x-noformat → Dictation en-AU-x-noformat → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-GB-x-beta → Ambient en-GB-x-beta → Dictation en-GB-x-beta → Offline en-GB-x-noformat → Ambient en-GB-x-noformat → Dictation en-GB-x-noformat → Offline en-US-x-beta → Ambient en-US-x-beta → Dictation en-US-x-beta → Offline en-x-beta → Ambient en-x-beta → Dictation en-x-beta → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline en-x-noformat → Ambient en-x-noformat → Dictation en-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es dev-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline es-x-noformat → Ambient es-x-noformat → Dictation es-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr dev-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline fr-CH-x-beta → Ambient fr-CH-x-beta → Dictation fr-CH-x-beta → Offline fr-CH-x-noformat → Ambient fr-CH-x-noformat → Dictation fr-CH-x-noformat → Offline fr-x-beta → Ambient fr-x-beta → Dictation fr-x-beta → Offline fr-x-noformat → Ambient fr-x-noformat → Dictation fr-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual dev-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de dev-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en dev-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch dev-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-mock dev-par02/mlservices/asr-aed-mock	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no dev-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv dev-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
speechmatics-adapter dev-weu/base/speechmatics-adapter	0 External Speechmatics model. Language-agnostic adapter deployed per tenant.	Routing enabled via `ml-service-routing`	ar → Ambient ar-x-noformat → Ambient bg → Ambient bg-x-noformat → Ambient cmn → Ambient cmn-x-noformat → Ambient cs → Ambient cs-x-noformat → Ambient el → Ambient el-x-noformat → Ambient et → Ambient et-x-noformat → Ambient fi → Ambient fi-x-noformat → Ambient he → Ambient he-x-noformat → Ambient hi → Ambient hi-x-noformat → Ambient hr → Ambient hr-x-noformat → Ambient hu → Ambient hu-x-noformat → Ambient it → Ambient it-x-noformat → Ambient ja → Ambient ja-x-noformat → Ambient ko → Ambient ko-x-noformat → Ambient lt → Ambient lv → Ambient lv-x-noformat → Ambient mr → Ambient mr-x-noformat → Ambient ms → Ambient ms-x-noformat → Ambient mt → Ambient mt-x-noformat → Ambient nl → Ambient nl-x-noformat → Ambient no → Ambient no-x-noformat → Ambient pl → Ambient pl-x-noformat → Ambient pt → Ambient pt-x-noformat → Ambient ro → Ambient ro-x-noformat → Ambient ru → Ambient ru-x-noformat → Ambient sk → Ambient sk-x-noformat → Ambient sl → Ambient sl-x-noformat → Ambient sv → Ambient sv-x-noformat → Ambient sw → Ambient sw-x-noformat → Ambient ta → Ambient ta-x-noformat → Ambient th → Ambient th-x-noformat → Ambient tr → Ambient tr-x-noformat → Ambient uk → Ambient uk-x-noformat → Ambient ur → Ambient ur-x-noformat → Ambient vi → Ambient vi-x-noformat → Ambient yue → Ambient yue-x-noformat → Ambient	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input.

dev/base dev-par01/base Routing config 99 languages

Name	Generation	Status	Routing Keys	Modes
asr-tdt-da dev-par02/mlservices/asr-tdt-da	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	da → Ambient da → Dictation da → Offline da-x-beta → Ambient da-x-beta → Dictation da-x-beta → Offline da-x-noformat → Ambient da-x-noformat → Dictation da-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-de dev-par02/mlservices/asr-tdt-de	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	de → Ambient de → Dictation de → Offline de-CH → Ambient de-CH → Dictation de-CH → Offline de-CH-x-beta → Ambient de-CH-x-beta → Dictation de-CH-x-beta → Offline de-CH-x-noformat → Ambient de-CH-x-noformat → Dictation de-CH-x-noformat → Offline de-x-beta → Ambient de-x-beta → Dictation de-x-beta → Offline de-x-noformat → Ambient de-x-noformat → Dictation de-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-en dev-par02/mlservices/asr-tdt-en	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	en → Ambient en → Dictation en → Offline en-AU → Ambient en-AU → Dictation en-AU → Offline en-AU-x-beta → Ambient en-AU-x-beta → Dictation en-AU-x-beta → Offline en-AU-x-noformat → Ambient en-AU-x-noformat → Dictation en-AU-x-noformat → Offline en-GB → Ambient en-GB → Dictation en-GB → Offline en-GB-x-beta → Ambient en-GB-x-beta → Dictation en-GB-x-beta → Offline en-GB-x-noformat → Ambient en-GB-x-noformat → Dictation en-GB-x-noformat → Offline en-US-x-beta → Ambient en-US-x-beta → Dictation en-US-x-beta → Offline en-x-beta → Ambient en-x-beta → Dictation en-x-beta → Offline en-x-multi → Ambient en-x-multi → Dictation en-x-multi → Offline en-x-noformat → Ambient en-x-noformat → Dictation en-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-es dev-par02/mlservices/asr-tdt-es	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	es → Ambient es → Dictation es → Offline es-x-noformat → Ambient es-x-noformat → Dictation es-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-fr dev-par02/mlservices/asr-tdt-fr	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing enabled via `ml-service-routing`	fr → Ambient fr → Dictation fr → Offline fr-CH → Ambient fr-CH → Dictation fr-CH → Offline fr-CH-x-beta → Ambient fr-CH-x-beta → Dictation fr-CH-x-beta → Offline fr-CH-x-noformat → Ambient fr-CH-x-noformat → Dictation fr-CH-x-noformat → Offline fr-x-beta → Ambient fr-x-beta → Dictation fr-x-beta → Offline fr-x-noformat → Ambient fr-x-noformat → Dictation fr-x-noformat → Offline	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-tdt-multilingual dev-par02/mlservices/asr-tdt-multilingual	3 TDT-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	Routing not configured via `ml-service-routing`	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-de dev-par02/mlservices/asr-aed-de	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-en dev-par02/mlservices/asr-aed-en	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-gswch dev-par02/mlservices/asr-aed-gswch	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-mock dev-par02/mlservices/asr-aed-mock	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-no dev-par02/mlservices/asr-aed-no	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
asr-aed-sv dev-par02/mlservices/asr-aed-sv	2 AED-based model on Triton via gRPC request/response. Shared across tenants with dynamic routing to inference clusters.	$ AED model with no routing. Traffic for this language is routed to TDT. Potential cost savings by decommissioning this model.	—	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input. Offline Batch transcription of pre-recorded audio. Processes complete audio files asynchronously.
speechmatics-adapter dev/base/speechmatics-adapter	0 External Speechmatics model. Language-agnostic adapter deployed per tenant.	Routing enabled via `ml-service-routing`	ar → Ambient ar-x-noformat → Ambient bg → Ambient bg-x-noformat → Ambient cmn → Ambient cmn-x-noformat → Ambient cs → Ambient cs-x-noformat → Ambient el → Ambient el-x-noformat → Ambient et → Ambient et-x-noformat → Ambient fi → Ambient fi-x-noformat → Ambient he → Ambient he-x-noformat → Ambient hi → Ambient hi-x-noformat → Ambient hr → Ambient hr-x-noformat → Ambient hu → Ambient hu-x-noformat → Ambient it → Ambient it-x-noformat → Ambient ja → Ambient ja-x-noformat → Ambient ko → Ambient ko-x-noformat → Ambient lt → Ambient lv → Ambient lv-x-noformat → Ambient mr → Ambient mr-x-noformat → Ambient ms → Ambient ms-x-noformat → Ambient mt → Ambient mt-x-noformat → Ambient nl → Ambient nl-x-noformat → Ambient no → Ambient no-x-noformat → Ambient pl → Ambient pl-x-noformat → Ambient pt → Ambient pt-x-noformat → Ambient ro → Ambient ro-x-noformat → Ambient ru → Ambient ru-x-noformat → Ambient sk → Ambient sk-x-noformat → Ambient sl → Ambient sl-x-noformat → Ambient sv → Ambient sv-x-noformat → Ambient sw → Ambient sw-x-noformat → Ambient ta → Ambient ta-x-noformat → Ambient th → Ambient th-x-noformat → Ambient tr → Ambient tr-x-noformat → Ambient uk → Ambient uk-x-noformat → Ambient ur → Ambient ur-x-noformat → Ambient vi → Ambient vi-x-noformat → Ambient yue → Ambient yue-x-noformat → Ambient	Ambient Real-time streaming transcription for ambient listening. Processes audio continuously as it arrives. Dictation Real-time streaming transcription for dictation. Optimized for single-speaker direct speech input.