By No Code MBA · Facts checked
- Provider
- AssemblyAI
- Announcement date
- Status at last check
- Generally available
What changed
AssemblyAI-STT-3.7 is the recorded-audio model in the new STT 3.7 family and replaces Universal-3.5 Pro for pre-recorded transcription. AssemblyAI says it supports 32 languages with automatic detection and an optional 99-language fallback.
Access and availability
Use the AssemblyAI Pre-recorded Speech-to-Text API and select the model ID assemblyai-stt-3-7.
Official access documentationWhat you can build
Test transcription on representative recordings with domain vocabulary, then review names and numbers before using the output in downstream workflows.
Limitations
This model is for recorded audio, not live streaming. The optional 99-language fallback is distinct from the 32 languages available with automatic detection.