Skip to main content
Glama
Sign In
Routara MCP Server

Routara MCP Server

by 36412749-collab
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Remote

Routara MCP Server

npm version MCP Registry License: MIT

Official Model Context Protocol server for Routara. Use Routara chat, image, and video models from Cursor, Claude Desktop, Codex, Windsurf, VS Code, and other MCP clients.

Practical walkthroughs: Python SDK · Node.js SDK · Cursor MCP · Claude Desktop MCP

Install

npx -y routara-mcp

Or install the one-click routara-mcp.mcpb bundle from the latest GitHub release.

Related MCP server: genpark-multi-model-cost-latency-router-skill

Tools

Tool

Description

routara_list_models

Search and paginate the live model catalog

routara_chat

Single-turn or multi-turn chat, including tool calls

routara_generate_image

Text-to-image and reference-image generation

routara_generate_video

Text-to-video and image-to-video submission

routara_get_video_status

Poll an asynchronous video task

MCP client configuration

{
  "mcpServers": {
    "routara": {
      "command": "npx",
      "args": ["-y", "routara-mcp"],
      "env": {
        "ROUTARA_API_KEY": "sk-or-v1-YOUR_KEY_HERE"
      }
    }
  }
}

OpenAI Codex

[mcp_servers.routara]
command = "npx"
args = ["-y", "routara-mcp"]
enabled = true

[mcp_servers.routara.env]
ROUTARA_API_KEY = "sk-or-v1-YOUR_KEY_HERE"

After adding the config, restart your MCP client and ask it to list Routara models. Then ask it to use routara_chat with a model available to your account. The MCP connection exposes tools; it does not replace the host application's own model subscription or billing.

Environment

Variable

Required

Default

ROUTARA_API_KEY

Yes for tool calls

None

ROUTARA_API_BASE

No

https://api.routara.ai/v1

ROUTARA_API_TIMEOUT_MS

No

30000

The process can start and complete the MCP handshake without an API key. A key is resolved only when a Routara tool is invoked, which lets directories validate the server safely.

Reliability and compatibility

  • Requests time out instead of hanging indefinitely.

  • Transient network errors, HTTP 429, and HTTP 5xx responses are retried with bounded backoff.

  • API errors include the upstream request ID and retry delay when available.

  • routara_chat returns the complete OpenAI-compatible response, preserving reasoning content and tool calls.

  • Media parameters are forwarded only when explicitly supplied; model-specific support varies.

Development

npm ci
npm run build
npm test
npm run validate:registry
npm run build:mcpb

For a live API smoke test:

ROUTARA_API_KEY=sk-or-v1-... npm run test:live

Directory listings

Security

Do not commit API keys. If a key is exposed, revoke it in the Routara dashboard immediately. Please report security issues privately to support@routara.ai.

License

MIT

Available Tools

5 tools
routara_chatA

Create an OpenAI-compatible chat completion through Routara.

Use message for a simple single-turn request, or messages for system prompts, multi-turn chat, multimodal content, and tool-call continuations. The complete upstream response is returned, including usage, reasoning_content, and tool_calls when supplied by the model.

Inputs: model (required), stop, tools, top_p, +5 more.

routara_generate_imageB

Generate or edit an image through Routara.

Model support varies. Besides prompt and size, the tool can pass quality, aspect ratio, reference image URL, negative prompt, seed, style, and response format to compatible models.

Inputs: model (required), prompt (required), n, seed, +7 more.

routara_generate_videoA

Submit an asynchronous video-generation job through Routara.

Supports text-to-video and image-to-video. Save the returned id or task_id and poll it with routara_get_video_status until completed or failed.

Inputs: model (required), prompt (required), fps, duration, +4 more.

routara_get_video_statusB

Poll a Routara video-generation job.

Pass the model when the upstream provider requires model-specific status routing.

Inputs: task_id (required), model.

routara_list_modelsA

List models available through Routara.

Use this before another tool when a model was not specified. Supports pagination and text filtering. The count field is always the total number of matches before pagination; returned is the number included in this response.

Inputs: limit, query, offset.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Deploy Server
    One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

    TDQS

    A3.9/5.0

    Scored across 5 tools

    Disambiguation5/5

    Each tool targets a distinct operation (listing models, chat, image generation, video submission, status polling) with no overlapping functionality. An agent can clearly distinguish between them.

    Naming Consistency4/5

    All tools are prefixed with 'routara_' and follow a verb_noun pattern (list_models, chat, generate_image, generate_video, get_video_status). The only minor inconsistency is 'chat' not using an underscore, but it remains clear.

    Tool Count5/5

    5 tools is an appropriate size for a server offering AI model interaction services. It is neither sparse nor overwhelming, covering the essential operations.

    Completeness4/5

    The set covers main workflows: listing models, chat completions, image generation, video submission, and status polling. A potential gap is lack of a cancel operation for video jobs, but core functionality is present.

    Maintenance

    ActivityNo data
    ResponsivenessOutdated data

    Resources

    Related MCP Connectors

    Related MCP Servers

    • farebox-mcp

      fareboxfun
      A
      license
      A
      quality
      C
      maintenance
      Connects Claude Desktop, Cursor, or any MCP-compatible client to 36+ frontier LLMs with automatic per-token payments in USDC on Solana.
      5
      17 npm
      MIT
    • genpark-multi-model-cost-latency-router-skill

      genpark-multi-model-costofficial

      alphaparkinc
      F
      license
      Not graded
      quality
      B
      maintenance
      Enables AI agents and MCP clients to dispatch queries to the most suitable model among 70+ LLMs by scoring intent, budget, cost, and latency, returning structured routing decisions with execution telemetry. It runs as a zero-dependency Python MCP server that plugs into Claude Desktop, Cursor, and other MCP-compatible clients.
      8
      -
    • genpark-multi-model-cost-latency-router-skill

      genpark-multi-model-costofficial

      Alpha-Park
      F
      license
      Not graded
      quality
      Not graded
      maintenance
      Enables dynamic, intent- and budget-aware routing of queries across 70+ LLMs by balancing cost and latency through Pareto-optimal model selection. Integrates with MCP clients like Claude Desktop and Cursor for automated multi-model execution.
      -
    • zerouter-mcp

      zerouter-mcpofficial

      ZerouterAI
      A
      license
      A
      quality
      B
      maintenance
      Connects MCP clients such as Claude Desktop, Cursor and Claude Code to an OpenAI- and Anthropic-compatible AI gateway, enabling chat with Claude, GPT, Gemini, Grok and Kimi models, listing available models, and looking up per-million-token pricing. It also sends one prompt to several models in parallel and returns each answer with token usage, latency and estimated cost.
      4
      MIT