The Voice MCP server provides pay-per-call speech synthesis via the https://voice.forgemesh.io API, supporting 20 voices, 10 personas, and 31 languages. Payments are made in USDC on the Base network; no API keys or subscriptions are needed.

• list_voice_catalog (Free): Discover all available voices, personas, language codes, pricing tiers, character limits, and speed/quality controls before making any paid call.

• generate_standard_voice ($0.001–$0.003): Convert text (up to 2,000 characters) to WAV audio using 10 standard voices (M1–M5, F1–F5) across 31 languages. Ideal for simple narration, alerts, and status updates.

• generate_controlled_voice ($0.003–$0.006): Produce WAV speech with granular control over speed (0.7x–2.0x) and quality steps (1–100) for precise pacing or polished audio fidelity.

• generate_persona_voice ($0.005–$0.01): Synthesize expressive WAV audio using 10 distinct persona voices (e.g., Storyteller, Narrator, Announcer, Assistant, Urgent, Velvet, Echo), with speed and quality controls. Best for branded agents, characters, and premium experiences.

• generate_openai_compatible_voice ($0.001–$0.003): Submit speech requests using an OpenAI /v1/audio/speech-shaped payload (input, voice, model, response_format), supporting WAV, FLAC, and OGG output. Ideal for apps already built around the OpenAI audio API format.

• generate_batch_voices ($0.002–$0.005): Process up to 20 texts in a single call with per-item voice and language settings, perfect for notification queues, scripted sequences, and multi-step agent workflows.

All audio is returned as audio_base64-encoded output. Input validation (voice names, language codes, character limits, speed/quality ranges) happens locally before any paid call is made.

Provides tools for generating speech using an OpenAI-compatible API, enabling text-to-speech with multiple voices, personas, languages, speed, quality controls, and batch generation.

1. Click on "Install Server".

2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.

3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Voice MCP speak 'Hello, how are you?' in F1 voice"

That's it! The server will respond to your query, and you can continue using it as needed.

Here is a step-by-step guide with screenshots.