# Patch AMD

> Patch AMD is an answering machine detection API for outbound calls. Stream the call audio from Twilio, Telnyx or any dialer and get one decision — human, voicemail, IVR, unavailable or silence — usually within a second of the greeting.

Price: $0.007 per answered detection (volume tiers to $0.004; first 250 a month free; detector failures free). Results: human, voicemail, ivr, unavailable, silence, not_sure. OpenAMD, English subset, 128 clips: 96.9% accuracy, 0 recordings connected.

## Docs

- [Quickstart](https://patchamd.com/docs.md): Stream a call's audio to Patch AMD and get a human-or-machine decision in under ten minutes.
- [Twilio guide](https://patchamd.com/docs/twilio.md): Use Patch AMD with Twilio Media Streams instead of Twilio's built-in answering machine detection.
- [Telnyx guide](https://patchamd.com/docs/telnyx.md): Use Patch AMD with Telnyx media streaming instead of Telnyx answering machine detection.
- [Raw WebSocket guide](https://patchamd.com/docs/raw-websocket.md): Stream raw mu-law, A-law or PCM audio from any dialer or media server to Patch AMD.
- [Webhooks](https://patchamd.com/docs/webhooks.md): Receive and verify the signed amd.decided webhook.
- [API reference](https://patchamd.com/docs/api-reference.md): Patch AMD REST endpoints, the stream WebSocket protocol, results and reason codes.

## Specs

- [OpenAPI 3.1](https://patchamd.com/openapi.json): REST endpoints (mint stream URLs, read detections, usage)
- [AsyncAPI 3.0](https://patchamd.com/asyncapi.json): the /v1/stream WebSocket protocol

## Optional

- [Full docs in one file](https://patchamd.com/llms-full.txt)
- [Pricing](https://patchamd.com/pricing)
- [Benchmark](https://patchamd.com/benchmark)
- [Privacy](https://patchamd.com/privacy)