Updated

AssemblyAI-STT-3.7-Realtime-Agent

AssemblyAI released AssemblyAI-STT-3.7-Realtime-Agent for voice-agent conversations. It supports 32 languages with automatic detection and is designed to focus transcription on the person speaking to the agent.

By No Code MBA · Facts checked

Provider
AssemblyAI
Announcement date
Status at last check
Generally available

What changed

AssemblyAI-STT-3.7-Realtime-Agent is the voice-agent model in the STT 3.7 family and replaces Universal-3.6 Pro Realtime. AssemblyAI says it reduces background speech and supports 32 languages in one stream with automatic detection.

Access and availability

Use the AssemblyAI Realtime Speech-to-Text API and set speech_model to assemblyai-stt-3-7-realtime-agent.

Official access documentation

What you can build

Evaluate it on calls with background speech and short numeric responses, checking transcripts against consented test audio before deployment.

Limitations

AssemblyAI recommends near-field voice focus in some environments where another speaker is close to the microphone. Reported quality and latency figures are company evaluations.

Sources and release history