[FEATURE]: Voice input MCP server for CLI terminal tools #41413

Activity

1. github-actions

   assigned

   jlongster

   on Aug 9, 2026
2. github-actions

   added

   needs:complianceThis means the issue will auto-close after 2 hours.This means the issue will auto-close after 2 hours.

   on Aug 9, 2026
3. github-actions commented on Aug 9, 2026

   

   github-actionsBot

   on Aug 9, 2026 – with GitHub Actions

   Contributor

   More actions

   This issue does not fully meet our contributing guidelines.

   What needs to be fixed:

   • Feature requests must use the **[FEATURE]: ** prefix in the title (e.g. [FEATURE]: Add voice input support via MCP server)
   • The feature request template requires a verification checkbox confirming you have searched for existing issues — this was not included

   Please edit this issue to address the above within 2 hours, or it will be automatically closed.

   ---

   This issue might be a duplicate of existing issues. Please check:

   • [FEATURE] Desktop: voice input for prompts (mic → text) and optional voice summary output #41364: [FEATURE] Desktop: voice input for prompts (mic → text) and optional voice summary output
   • [FEATURE]: Voice/Speech Input Support for Desktop App #33300: [FEATURE]: Voice/Speech Input Support for Desktop App
   • Feature: Hold-spacebar push-to-talk voice input (like Claude Code) #35219: Feature: Hold-spacebar push-to-talk voice input (like Claude Code)
   • [FEATURE]: Clickable microphone button in OpenCode Desktop app #37742: [FEATURE]: Clickable microphone button in OpenCode Desktop app
   • [Feature Request] Entrada por microfono / reconocimiento de voz #39688: [Feature Request] Entrada por microfono / reconocimiento de voz (voice/microphone input)

   If you believe this was flagged incorrectly, please let a maintainer know.
4. YuvrajSinghBhadoria2

   changed the title [-]feat: Add voice input support via MCP server[/-] [+][FEATURE]: Voice input MCP server for CLI terminal tools[/+] on Aug 9, 2026
5. github-actions

   removed

   needs:complianceThis means the issue will auto-close after 2 hours.This means the issue will auto-close after 2 hours.

   on Aug 9, 2026
6. github-actions commented on Aug 9, 2026

   

   github-actionsBot

   on Aug 9, 2026 – with GitHub Actions

   Contributor

   More actions

   Thanks for updating the issue! It now meets the contributing guidelines. We appreciate the detailed write-up and the effort to differentiate this from existing voice input requests.
7. YuvrajSinghBhadoria2 commented on Aug 15, 2026

   

   YuvrajSinghBhadoria2

   on Aug 15, 2026

   Author

   More actions

   Update: implementation published 🎤

   The voice MCP server is built and published to npm. Note: it's published under my own scope (@yuvarjbhado) since I don't own the @opencode-ai org on npm — the install snippet in the original post should use this name:

   {
     "mcpServers": {
       "voice": {
         "command": "npx",
         "args": ["-y", "@yuvarjbhado/voice-mcp"]
       }
     }
   }

   Or with OpenCode's mcp key:

   {
     "mcp": {
       "voice": {
         "command": "npx",
         "args": ["-y", "@yuvarjbhado/voice-mcp"]
       }
     }
   }

   What it provides:

   • voice_transcribe — record + local Whisper transcription
   • voice_type — record, transcribe, and type at cursor
   • voice_status — check availability

   100% local (faster-whisper), cross-platform (sox/ffmpeg), no cloud API. Source: https://github.com/YuvrajSinghBhadoria2/opencode-voice-mcp

   Still happy to fold this into the opencode repo (docs or a packages/ dir) if maintainers want it core — just let me know the preferred location.
8. x-lixu commented on Aug 18, 2026

   

   x-lixu

   on Aug 18, 2026

   Last edited by x-lixu

   More actions

   Nice work shipping the dictation MCP — push-to-talk transcription is a solid baseline for terminal tools.

   Worth noting there's a complementary approach if you want conversation rather than dictation: qwen-audio-agent (disclosure: I maintain it) runs a full-duplex voice frontend in front of CLI agents including OpenCode:

   • No push-to-talk: talk while the agent works, interrupt mid-task (barge-in)
   • Spoken results: long-running task outcomes are announced back by voice, so you can step away from the terminal
   • Voice permission gates: destructive tool calls require spoken confirmation
   • Local wake word (sherpa-onnx, on-device)

   OpenCode is a first-class backend — qwenaudio install opencode sets it up, then qwenaudio --backend opencode (or the desktop orb) starts a voice session over ACP.

   The two aren't mutually exclusive: a dictation MCP covers "type by voice" inside the TUI, while a runtime like this covers "supervise the agent hands-free".
9. YuvrajSinghBhadoria2 commented on Sep 6, 2026

   

   YuvrajSinghBhadoria2

   on Sep 6, 2026

   Author

   More actions

   @jlongster — wanted to check in on this one since it's been quiet for a few weeks. The voice MCP server is live and published (@yuvarjbhado/voice-mcp, 100% local faster-whisper). The earlier docs PR (#42769) was raised as official-docs placement, but I'd rather match what maintainers actually want than push a personal package into the Guides. Two options:

   1. Keep voice input core/documented inside opencode (docs + optionally a packages/ integration), or
   2. List it in the ecosystem listing instead.

   Happy to fold the guide into whichever home you prefer — or close this out cleanly if core voice input isn't on the roadmap. Thanks!
10. neriousy commented on Oct 20, 2026

    

    neriousy

    on Oct 20, 2026

    Member

    More actions

    Thanks for building and publishing this voice MCP server. We're closing the proposal to include it in OpenCode.
11. neriousy

    closed this as not plannedon Oct 20, 2026