Repolode
Submit
dograh-hq

dograh-hq/dograh

Self-hostable open-source voice AI platform for building, testing and deploying phone and browser voice agents, with a visual workflow builder and MCP support.

Open-source alternative to:VapiRetell

View on GitHubWebsite
Stars
5,906
Forks
1,508
Open issues
99
Last commit
yesterday

What is it?

Dograh is an open-source platform for building voice agents, positioned as a self-hostable alternative to commercial services such as Vapi and Retell. Developers and teams use its visual workflow builder to define call flows with agent nodes, transitions, tools and end-call outcomes, then test the agent in the browser by voice or by text chat before connecting it to a phone number.

The platform supports both inbound and outbound calls and includes telephony integrations for Twilio, Vonage, Telnyx, Plivo and Asterisk ARI. Users can bring their own LLM, speech-to-text and text-to-speech providers, or use the bundled stack. Artifacts can be stored in the bundled MinIO or in AWS/S3-compatible storage.

Dograh runs on Docker through a single setup script and includes a Python backend, Python and Node SDKs, and an MCP server that lets coding assistants inspect agents and create or edit workflow drafts.

Key features

  • 01Visual workflow builder with agent nodes, tools, transitions and end-call outcomes
  • 02Browser testing by voice (Test Audio) or text chat (Test Chat)
  • 03Inbound and outbound calling with Twilio, Vonage, Telnyx, Plivo and Asterisk ARI integrations
  • 04Bring-your-own LLM, STT and TTS providers, or use the bundled stack
  • 05MCP server for coding assistants to create and edit agent workflows
  • 06Python and Node SDKs for programmatic agent creation and outbound calls

Use cases

  • Building lead qualification or appointment-setting phone bots
  • Running outbound calling campaigns driven by a programmatic SDK
  • Self-hosting voice agents where call data must stay on your own infrastructure
  • Letting coding agents draft and revise voice agent workflows through MCP

Quick start

  1. 1

    Download the compose file and startup script, make the script executable, and run it

    curl -o docker-compose.yaml https://raw.githubusercontent.com/dograh-hq/dograh/main/docker-compose.yaml && curl -o start_docker.sh https://raw.githubusercontent.com/dograh-hq/dograh/main/scripts/start_docker.sh && chmod +x start_docker.sh && ./start_docker.sh
  2. 2

    Wait 2-3 minutes for images to download, then open the app in your browser

  3. 3

    Open http://localhost:3010 and create your first voice bot

Things to know

  • The quickstart starts a public trycloudflare.com tunnel and enables signup by default, so you should disable signup after creating your account.
  • Connecting your own telephony or LLM/TTS/STT providers requires separate API keys and accounts.

Topics

#voice-ai#voice-agents#telephony#conversational-ai#self-hosted#speech-to-text#text-to-speech#mcp#ai-calling#asterisk-ari

FAQ

What is dograh?
Self-hostable open-source voice AI platform for building, testing and deploying phone and browser voice agents, with a visual workflow builder and MCP support. Dograh is an open-source platform for building voice agents, positioned as a self-hostable alternative to commercial services such as Vapi and Retell. Developers and teams use its visual workflow builder to define call flows with agent nodes, transitions, tools and end-call outcomes, then test the agent in the browser by voice or by text chat before connecting it to a phone number.
Is dograh free to use? What license does it have?
dograh is open source under the BSD-2-Clause license. Check the license terms before using it in commercial products.
Is dograh still maintained?
Actively maintained. The last commit to dograh was on Oct 19, 2026.
What can I use dograh for?
Building lead qualification or appointment-setting phone bots Running outbound calling campaigns driven by a programmatic SDK Self-hosting voice agents where call data must stay on your own infrastructure Letting coding agents draft and revise voice agent workflows through MCP
How do I install dograh?
Download the compose file and startup script, make the script executable, and run it: curl -o docker-compose.yaml https://raw.githubusercontent.com/dograh-hq/dograh/main/docker-compose.yaml && curl -o start_docker.sh https://raw.githubusercontent.com/dograh-hq/dograh/main/scripts/start_docker.sh && chmod +x start_docker.sh && ./start_docker.sh Wait 2-3 minutes for images to download, then open the app in your browser Open http://localhost:3010 and create your first voice bot
Where can I get dograh?
The source code and installation instructions are on GitHub: https://github.com/dograh-hq/dograh

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects