Omni runs the whole call. The rest are building blocks.
Omni is one speech-to-speech model for telephony agents - no stitching STT, an LLM, turn detection, and TTS together. Agents, Trace, and Cast carry the rest of production voice work; Hear, Speak, and Telephony are the individual APIs underneath. Start free with $50 in credits.
Omni
The definitive all-in-one AI voice agent model. Hybrid speech-to-speech, fused LLM brain, emotion-aware voices, and tool calling.
Agents
Manage, evaluate & ship AI voice agents. No code, no engineering.
Cast
Cinematic, emotional long-form TTS for podcasts, narration, and audiobooks.
Trace
Bulletproof compliance & QA on every single call, automatically.
APIs and add-ons
Hear
Hear - speech-to-text (STT), telephony-native.
Speak
Speak - text-to-speech (TTS) that starts in milliseconds.
Cue
Telephony-grade hearing & turn detection - bring your own LLM and voice.
Telephony
One agent stack across your carriers and phone numbers.
AMD
Know who answered. Even when it's an AI screening the call.
Recap
Transform every call into instant, structured intelligence.
| Model | What it does | Endpoint | Price | Status |
|---|---|---|---|---|
| Omni | The definitive all-in-one AI voice agent model. Hybrid speech-to-speech, fused LLM brain, emotion-aware voices, and tool calling. | wss://api.pyai.com/v1/omni | $0.05/min | Live |
| Agents | Manage, evaluate & ship AI voice agents. No code, no engineering. | Console + wss://api.pyai.com/v1/omni | $0.08/min | Coming soon |
| Cast | Cinematic, emotional long-form TTS for podcasts, narration, and audiobooks. | POST /v1/audio/speech | $0.02/min | Live |
| Trace | Bulletproof compliance & QA on every single call, automatically. | GET /v1/trace/interactions | $0.03/min | Live |
| Hear | Hear - speech-to-text (STT), telephony-native. | POST /v1/audio/transcriptions | $0.001/min | Live |
| Speak | Speak - text-to-speech (TTS) that starts in milliseconds. | POST /v1/audio/speech | $0.04/min | Live |
| Cue | Telephony-grade hearing & turn detection - bring your own LLM and voice. | wss://api.pyai.com/v1/audio/transcriptions/stream | $0.015/min | Live |
| Telephony | One agent stack across your carriers and phone numbers. | POST /v1/telephony/numbers | $0.01/min | Beta |
| AMD | Know who answered. Even when it's an AI screening the call. | wss://api.pyai.com/v1/amd/stream | $0.004/min | Live |
| Recap | Transform every call into instant, structured intelligence. | GET /v1/recap/calls | $0.02/min | Add-on |
Ship phone agents on one speech-to-speech model.
Start with a free key and $50 in credits. Build on the same OpenAI-compatible stack you take to production with Omni, Agents, and Trace - natural turn-taking, telephony-native, live in an afternoon.
No credit card - OpenAI-compatible - cancel anytime