Openrouter Multimodal MCP Server

OpenRouter MCP Multimodal The MCP server for multimodal AI agents. One install · 19 tools · 300+ OpenRouter models · text, vision, audio & video - analysis and generation. It runs locally over stdio via the published package.

People connecting cloud infrastructure tools to Claude, Cursor, VS Code, or another MCP client. The project is written in TypeScript.

VERIFIED ACTIVE

LAST COMMIT 2026-08-25 · ★ 81 · #8 OF 141 MAINTAINED CLOUD INFRASTRUCTURE · VERIFIED 2026-08-25

Apache-2.0 · TypeScript servers · how we verify → /methodology

01 · Install Openrouter Multimodal

before you install - you'll need

Set OPENROUTER_API_KEY before connecting.

Claude Code

claude mcp add stabgan-openrouter-mcp-multimodal -- npx -y @stabgan/openrouter-mcp-multimodal

Claude Desktop / Cursor / VS Code - add to config

{
  "mcpServers": {
    "stabgan-openrouter-mcp-multimodal": {
      "command": "npx",
      "args": [
        "-y",
        "@stabgan/openrouter-mcp-multimodal"
      ]
    }
  }
}

Same JSON for Cursor. For VS Code, rename the top-level key from `mcpServers` to `servers`.

Using another client? Same JSON, different key

Claude Desktop · mcpServers

Cursor · mcpServers

VS Code · servers

Windsurf · mcpServers

Zed · context_servers

Cline · mcpServers

Roo Code · mcpServers

Continue · mcpServers

LibreChat · mcpServers

Gemini CLI · mcpServers

Codex CLI · mcp_servers

Full setup guides: every client.

02 · Evidence

Security posture

What to check before giving this server access to your agent - from the registry, GitHub, and our own probes. We don't score safety; we show what's verifiable.

runs as local process (stdio) - runs on your machine with your user's permissions · container (OCI image) - can be sandboxed away from your filesystem

license Apache-2.0 - declared in the repository

npm package @stabgan/openrouter-mcp-multimodal - published under the repo owner's npm scope (@stabgan)

registry namespace io.github.stabgan is GitHub-verified and matches the repo owner

03 · What Openrouter Multimodal can do

Prose above is summarized from the project's README and registry record - no invented capabilities.

What you can build

An agent gets 19 documented tools spanning generate, image, video, chat, including chat_completion, start_chat_completion, get_chat_completion_status, analyze_image.

Capability map

Tools grouped from the project's README - what Openrouter Multimodal lets an agent do.

Generate

5 tools - e.g. Text-to-image via chat completions with reference images; Text-to-image via dedicated /api/v1/images (resolution, quality, format); Text-to-speech / music via chat completions

generate_image · generate_image_dedicated · generate_audio · generate_video · generate_video_from_image

Chat

3 tools - e.g. Text chat, web search, provider routing, caching, reasoning; Async background job for long-running reasoning models; Poll / retrieve async completion results

chat_completion · start_chat_completion · get_chat_completion_status

Analyze

3 tools - e.g. Vision - local path, URL, or data URL + question; Transcribe / analyze audio files; Describe / Q&A over video files

analyze_image · analyze_audio · analyze_video

read & search

3 tools - e.g. Poll / resume video jobs; Paginated model catalog search; Pricing, context, modalities

get_video_status · search_models · get_model_info

Text

1 tool - e.g. Dedicated TTS (/api/v1/audio/speech) - voices, speed, format

text_to_speech

Speech

1 tool - e.g. Dedicated STT (/api/v1/audio/transcriptions) - Whisper, GPT-4o

speech_to_text

Validate

1 tool - e.g. Cheap model ID existence check

validate_model

Rerank

1 tool - e.g. Relevance ranking for RAG

rerank_documents

Latest releases

v4.6.1 · 2026-08-22

Adds SVG icons for all 19 tools and server metadata (title, description, websiteUrl, icon) per MCP 2025-11-25 spec. Purely additive - no breaking changes. · npx @stabgan/[email protected]

v4.6.0 · 2026-08-22

What's New · Wrap frame_images/input_references in {type: "image_url", image_url: {url}} format - fixes Zod validation errors on image-to-video (closes #23) · Detect non-writable Windows system directories…

v4.5.1 · 2026-05-04

Patch release fixing 2 HIGH + 4 MEDIUM findings from an independent post-release audit. Fully backwards compatible with v4.5.0. · Fixed (HIGH): · MCP initialize handshake advertised stale 4.0.1 - now uses…

04 · Who maintains Openrouter Multimodal

openrouter-multimodal is maintained by stabgan. We track 2 MCP servers from stabgan - 2 actively maintained, 82 combined GitHub stars, oldest repo from Mar 2025.

05 · Facts

category
cloud infrastructure - ranked #8 of 141 actively-maintained cloud infrastructure servers as of 2026-08-25.
release cadence
2 releases in the last 90 days (latest 2026-08-22)
registry
io.github.stabgan/openrouter-multimodal (active, first published 2026-05-03 · 5 versions)
packages
npm:@stabgan/openrouter-mcp-multimodal · oci:docker.io/stabgan/openrouter-mcp-multimodal:4.5.1

06 · Openrouter Multimodal FAQ

Is Openrouter Multimodal still maintained?

Yes - as of 2026-08-25, its last commit was 2026-08-25 and it shipped 2 releases in the last 90 days. We re-verify nightly.

What can Openrouter Multimodal do?

An agent gets 19 documented tools spanning generate, image, video, chat, including chat_completion, start_chat_completion, get_chat_completion_status, analyze_image.

How do I install Openrouter Multimodal?

Run `npx -y @stabgan/openrouter-mcp-multimodal`. The README documents one environment variable (OPENROUTER_API_KEY) to set first. Set OPENROUTER_API_KEY before connecting. You can also paste the ready-made client config above.

Does Openrouter Multimodal run locally?

Yes - it's a stdio server: it runs on your machine (via npx) with your user's permissions. Your data stays local unless the server itself calls external APIs.

07 · Alternatives to Openrouter Multimodal

More cloud infrastructure MCP servers · Partd MCP · Shortcut MCP · Coolify · Emfirge · Costco

More TypeScript MCP servers · Starlog · Starwind UI MCP Server · Mintclub · see all