Skip to content

Developer toolkit

Natural voice. Your stack.

Build with Hear, Speak, or the complete Omni voice-agent loop. Pick your SDK and start with source you can read, copy, and run.

Getting started.mdAgent skill.mdWebhooks.md

Pick the SDK that fits your application

Every project includes an entry point, dependencies, environment template, and run instructions.

Python 3.10+ · v0.4.0

Python

Stream Speak and transcribe recordings with Hear through one reusable client.

Install the SDK
python -m pip install pyai-sdk==0.4.0
Open starter → Download Markdown

Node.js 22+ · v0.4.0

TypeScript

Use typed requests and native ReadableStream audio in your Node.js backend.

Install the SDK
npm install @pyai/sdk@0.4.0
Open starter → Download Markdown

Node.js 22+ · v0.4.2

Twilio

Bridge a Twilio Media Stream to Omni for the complete speech-to-speech loop.

Install the SDK
npm install @pyai/twilio@0.4.2
Open starter → Download Markdown

Python 3.10–3.13 · v0.1.2

LiveKit

Add Hear and Speak to an existing LiveKit Agents session.

Install the SDK
python -m pip install livekit-plugins-pyai==0.1.2
Open starter → Download Markdown

Python 3.10–3.13 · v0.1.2

Pipecat

Plug PyAI speech services into the transport and LLM pipeline you already use.

Install the SDK
python -m pip install pipecat-pyai==0.1.2
Open starter → Download Markdown

Start inside your coding agent

Use Cursor, Codex, or Claude Code with the same API contract and runnable starters.

Install Node.js 22+ and your coding agent first. Connect PyAI’s MCP server, then ask it to call get_started. For a sandbox test, ask it to call create_sandbox_key once. Keep live credentials in your environment.

Codex · terminal
codex mcp add pyai -- npx -y @pyai/mcp@0.2.0
Claude Code · terminal
claude mcp add --transport stdio --scope project pyai -- npx -y @pyai/mcp@0.2.0
Cursor · .cursor/mcp.json
{
  "mcpServers": {
    "pyai": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@pyai/mcp@0.2.0"
      ]
    }
  }
}

Merge the pyai entry into your existing mcpServers configuration. Enable it in Cursor’s MCP settings.

Paste into your coding agent
Read https://pyai.com/skill.md and https://pyai.com/sdks.md.
Use the matching starter to add PyAI voice to this project.
Keep credentials in environment variables. Run a short synthetic
audio test and report the result and any missing credentials.
Do not place a phone call or send a message unless I ask.

Want a reusable local skill? Copy the Codex and Claude Code skill install commands. Any tool that reads Markdown can use skill.md.

SDK questions

Which SDK should I choose?

Use Python or TypeScript for direct API work, Twilio for an Omni phone agent, or the LiveKit and Pipecat adapters when those frameworks already own your application loop.

Are the examples complete projects?

Yes. Downloads contain dependencies, an environment template, source, and run instructions. Each page lists exactly what the starter does and any extra accounts needed for a full voice agent.

Can my coding agent read the documentation?

Yes. Public Markdown includes the getting-started guide, skill, webhooks, and source for every SDK starter. The MCP server also exposes tools for API discovery and sandbox testing.

Will changing SDKs change speech quality?

The SDKs connect to the same PyAI services. Select the appropriate voice and language, then test representative audio and calls in your application.