Back to Apps

Multimodal MCP Client
by ejb503
Voice-powered AI interface combining Google Gemini's multimodality with MCP's extensible tooling system.
0 stars
Works in:claude
Exposes:ToolsResources
What it does
This client provides a modern, voice-first AI interface that integrates Google Gemini for multimodal understanding (text, voice, and visual inputs) with the Model Context Protocol (MCP) to execute complex agentic workflows.
Tools
Since this is a client that orchestrates other MCP servers, it doesn't expose tools itself but provides the infrastructure to run:
- Custom MCP Servers: User-configured servers via
mcp.config.custom.json. - Systemprompt MCP Servers: Integrated servers available via Systemprompt API key.
Installation
Local Setup
- Clone the repo:
git clone https://github.com/Ejb503/multimodal-mcp-client.git - Install:
npm install && cd proxy && npm install - Configure: Create
config/mcp.config.custom.jsonand add your API keys to.env. - Run:
npm run dev
Supported hosts
- claude
Choose your AI client and follow the steps below.
All clients
npm run dev





