Live transcription & captions
Words on screen as they're spoken.
Open a streaming Hear connection and render partials as they arrive - ideal for live captions, agent-assist, and realtime UIs. Move to Omni when you want the full turn-taking voice agent.
How it works
- 1
Stream audio up
Send pcm/audio frames over the streaming transcription socket.
- 2
Render partials
Display partial results immediately; replace with finals per utterance.
- 3
Move to full voice agents
Switch to Omni when endpointing, grounded context, speech output, and tools should run in one voice loop.
What this replaces
Manual coverage gaps
Add realtime captions or live transcripts with eager partials. That means fewer missed calls, fewer handoffs without context, and fewer hours spent catching up after the fact.
Stacked AI bills
Teams often pay separately for transcription, speech, realtime models, telephony, QA, and orchestration. PyAI keeps the core voice workflow on one usage-based stack.
Slow experimentation
Start with a free key, replay real conversations, compare latency and cost, then route production traffic only when the numbers and customer experience hold up.
Common questions
How does PyAI support live transcription & captions?
Open a streaming Hear connection and render partials as they arrive - ideal for live captions, agent-assist, and realtime UIs. Move to Omni when you want the full turn-taking voice agent.
What does it cost to test this use case?
Omni API is billed per minute for realtime voice agents. The Agents feature is available when you want to manage, evaluate & ship agents with no code: monitoring and call intelligence packaged for you. Every account starts with free credit, so teams can test real call flows before committing budget.
How does this reduce wasted AI spend?
PyAI reduces duplicated vendor bills, manual review work, and migration drag by putting transcription, speech, realtime agents, grounding, and call intelligence on one OpenAI-compatible stack.
How do we keep call quality high in production?
Ground the agent in your own content, bind tools for real actions, warm-transfer when needed, and use Trace or Recap when calls require QA, compliance review, summaries, and audit evidence.
Choose how you want to start.
Use an instant sandbox key for the API, or open Agents Beta and build without code. Sandbox usage never bills. Live keys require available credit.
Agents is live in beta. Sandbox keys have daily limits and never touch billing.