About this model or application
Universal-3.5 Pro and Universal-2 transcribe submitted recordings through POST /v2/transcript, billed on audio duration.
Family & other versions3 sources
Background for the family or other versions. These sources are counted separately from evidence for this release.
Family & other versions
Recorded-audio selectors and fallback between Universal-3.5 Pro and Universal-2.
Family & other versions
Recorded 3.5/2 audio-hour prices and realtime 3.6 session-hour base price, with extra feature tariffs.
Family & other versions
Recorded audio seconds and streaming WebSocket-open session duration are billed separately.
Model family catalog2
Current models
Official catalogUniversal-3.5 ProAvailableAPI
- Model ID
universal-3-5-pro- Released
- Not stated
- Availability
- Available
- Input
- Audio
- Output
- Text
- Deployment / access
- API
- Capabilities
- Transcription · Multilingual
First-party endpoint pricing
USD / audio hour- Input
- 0.21
Async POST /v2/transcript; submitted audio duration, prorated to seconds; paid add-ons extra.
Pre-recorded model: speech_models=[universal-3-5-pro] on POST /v2/transcript, USD 0.21/audio-hour. The same selector string is reused for a separate Streaming endpoint, with different billing. 18 languages; default fallback uses Universal-2.
Universal-2AvailableAPI
- Model ID
universal-2- Released
- Not stated
- Availability
- Available
- Input
- Audio
- Output
- Text
- Deployment / access
- API
- Capabilities
- Transcription · Multilingual
First-party endpoint pricing
USD / audio hour- Input
- 0.15
Async POST /v2/transcript; submitted audio duration, prorated to seconds; paid add-ons extra.
Pre-recorded model: speech_models=[universal-2] on POST /v2/transcript, USD 0.15/audio-hour; 99 languages.