Back
developers

VoxFlow CLI + MCP: AI Voice Tools for Developers

VoxFlow Team 2026-04-03 2 min read

VoxFlow isn't just a web app — it's a full developer toolkit. The voxflow CLI lets you synthesize speech, generate podcasts, transcribe audio, and translate videos directly from your terminal. The MCP server brings these capabilities into Claude Code and other AI assistants.

CLI: Quick Start

Install

npm install -g voxflow

Login

voxflow login

This opens your browser for secure OAuth login. Your token is cached locally at ~/.config/voxflow/token.json.

Your First Synthesis

voxflow say "Hello, this is my first AI voiceover" -o hello.wav

That's it. One command, one audio file.

CLI Commands Overview

Command What It Does
voxflow say "text" Synthesize text to speech
voxflow narrate script.txt Multi-segment narration from file
voxflow podcast "topic" Generate a full AI podcast
voxflow story "prompt" AI story with narration
voxflow dub subtitles.srt Dub video from subtitles
voxflow asr audio.mp3 Transcribe audio to text
voxflow translate subs.srt Translate subtitles
voxflow video-translate video.mp4 End-to-end: transcribe + translate + dub
voxflow voices Browse 200+ available voices
voxflow present "topic" Generate narrated video slides

Real-World Examples

Generate a podcast from a topic

voxflow podcast "The future of AI agents" \
  --voice "authoritative-boss,cutesy-girl" \
  --bgm jazz \
  -o ai-agents-podcast.wav

Narrate a blog post

cat blog-post.md | voxflow narrate --voice gentle-female -o narration.wav

Translate a video end-to-end

voxflow video-translate presentation.mp4 \
  --target-lang ja \
  -o presentation-japanese.mp4

Transcribe a meeting recording

voxflow asr meeting.mp3 --format srt -o meeting.srt

MCP Server: Use VoxFlow in Claude Code

VoxFlow provides an MCP (Model Context Protocol) server that integrates with Claude Code, enabling AI-powered voice synthesis directly in your development workflow.

Setup

Add VoxFlow to your Claude Code MCP configuration:

{
  "mcpServers": {
    "voxflow": {
      "command": "npx",
      "args": ["-y", "voxflow", "mcp"]
    }
  }
}

What You Can Do

Once configured, Claude Code can:

  • Synthesize speech — "Read this error message aloud"
  • Generate podcasts — "Create a podcast about this PR"
  • Search voices — "Find a Japanese female voice"
  • Transcribe audio — "Transcribe this recording"

Example Workflow

In Claude Code, you can say:

"Generate a 3-minute podcast explaining the changes in this PR, using a professional male voice"

Claude will use the VoxFlow MCP tools to analyze your code changes, write a script, and produce the audio — all without leaving your terminal.

API Access

For direct API integration, VoxFlow provides REST endpoints:

# Synthesize text
curl -X POST https://api.voxflow.studio/api/tts/synthesize-stream \
  -H "Authorization: Bearer $VOXFLOW_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"text": "Hello world", "voiceId": "v-female-R2s4N9qJ"}' \
  -o output.mp3

# List voices
curl https://api.voxflow.studio/api/tts/voices

API access is included on the free tier — 10,000 credits per month.

What's Next?