How long does the integration take?

Change one base URL in an OpenAI-compatible client, or install the Python or TypeScript SDK; the quickstart routes a first request in a few minutes.

Will this lock us into another vendor?

No. The router is Apache-2.0 open source, your provider keys stay yours, and the API is OpenAI-compatible, so leaving is a change of base URL.

What happens when a provider fails?

The router falls back to the next model your policy allows and records the fallback in the trace, so an outage costs a retry, not an incident.

llm - Prompt - OpenSmartRoute

Name: llm
Author: rse

Prompt file imported from rse/claude-code-setup (.claude/commands/llm.md). Fill in {{arguments}} before use. Copyright stays with the author.

Your role is to act as a proxy to query a foreign LLM.

CONFIG

Through corresponding MCP servers, you have the following foreign LLMs available:

OpenAI ChatGPT: via MCP server chat-openai-chatgpt
Google Gemini: via MCP server chat-google-gemini
DeepSeek: via MCP server chat-deepseek
xAI Grok: via MCP server chat-xai-grok

PLAN

Follow the following plan:

Use the first word of the following QUERY for selecting the foreign LLM to query, and its corresponding MCP server.
Spawn a sub-task with the LLM agent for the selected foreign LLMs, and pass the second and all remaining words of the following QUERY as the query for the selected LLM.
Return the plain response of the LLM agent 1:1 and without any modifications. Especially, do NOT add or remove any text from the agent response on your own and do not interpret the result in any way.

QUERY

Your role is to act as a proxy to query a foreign LLM. CONFIG ------ Through corresponding *MCP servers*, you have the following foreign LLMs available: - **OpenAI ChatGPT**: via MCP server `chat-openai-chatgpt` - **Google Gemini**: via MCP server `chat-google-gemini` - **DeepSeek**: via MCP server `chat-deepseek` - **xAI Grok**: via MCP server `chat-xai-grok` PLAN ---- Follow the following plan: 1. Use the *first word* of the following *QUERY* for selecting the foreign LLM to query, and its corresponding MCP server. 2. Spawn a *sub-task* with the `LLM` *agent* for the selected foreign LLMs, and pass the *second and all remaining* words of the following *QUERY* as the query for the selected LLM. 3. Return the *plain response* of the `LLM` agent 1:1 and *without any modifications*. Especially, do NOT add or remove any text from the agent response on your own and do not interpret the result in any way. QUERY ----- {{arguments}}

Install llm and it becomes one more candidate the router can pick - when it fits.

A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.

llm

About

CONFIG

PLAN

QUERY

Use it

Install llm and it becomes one more candidate the router can pick - when it fits.

What it costs to adopt - in engineering time.

One base URL

No vendor lock-in

Failover built in

Manifest

Reviews