You can connect Cursor’s MCP client to the ElevenLabs TTS service through a dedicated MCP HTTP endpoint. This enables you to synthesize speech on demand, manage voices, and receive real-time updates via the MCP integration. The backend serves REST, WebSocket, and an MCP server endpoint that Cursor can consume for seamless TTS interactions.

How to use

Connect Cursor to the Jessica TTS MCP server by adding a new MCP server with the following details. Name it something meaningful like Jessica TTS. Set Type to an HTTP MCP connection and use the provided endpoint URL. The MCP server endpoint you connect to is: http://localhost:9020/sse.

Once connected, you can trigger text-to-speech operations from Cursor using the available voice configurations and input text. You will be able to select a voice, request synthesis, and receive speech data or status updates through the MCP channel in real time. Use the MCP-driven interface to queue synthesis requests, switch voices, and monitor progress without leaving your Cursor workspace.

How to install

Prerequisites: you need Python 3.11+, Poetry for backend dependencies, Node.js 18+ for the frontend, and a Cursor MCP-capable environment.

# 1. Backend: clone and set up
git clone https://github.com/georgi-io/jessica.git
cd jessica

# 2. Backend: create Python virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# 3. Backend: install dependencies
poetry install

# 4. Backend: configure environment
cp .env.example .env
# Edit .env to insert your ElevenLabs API key

# 5. Backend: install pre-commit hooks
poetry run pre-commit install

# 6. Frontend: install dependencies
cd src/frontend
npm install

"

Run the backend and frontend development servers after configuring the environment.

Backend: activate env and start server

cd .. # back to project root if needed source .venv/bin/activate # On Windows: .venv\Scripts\activate python -m src.backend

Frontend: start dev server

cd src/frontend npm run dev

## Additional content

Configuration and runtime details include the following endpoints and environment notes.

Backend REST API: http://localhost:9020
WebSocket: ws://localhost:9020/ws
MCP Server (for Cursor): http://localhost:9020/sse
Frontend: http://localhost:5173

## Troubleshooting

Common issues include API key problems, connection issues to the MCP server, port conflicts, and WebSocket failures. Ensure the backend is running on the expected ports and that the ElevenLabs API key is correctly configured in the environment.

## Available tools

### tts\_api

TTS REST/WebSocket bridge and voice management for ElevenLabs integration, exposed via MCP to Cursor.