openrouter-models — Claude Code skill: what it does, how to install
Queries OpenRouter for 300+ AI models: pricing, context limits, provider latency and throughput.
npx skills add OpenRouterTeam/skills --skill openrouter-models -a claude-codeWhat openrouter-models does
The skill gives the agent access to the OpenRouter catalog through several TypeScript scripts. list-models.ts outputs all available models with sorting by recency, price, context size or throughput. search-models.ts finds models by name or modality (text, image, audio, file). resolve-model.ts turns a vague query like "claude sonnet" into an exact model identifier and returns a confidence score — at high (≥0.85) you can use it immediately, at medium you should confirm with the user. compare-models.ts lays out pricing and capabilities side by side, get-endpoints.ts shows latency (p50–p99), uptime and throughput for each provider — this script requires an API key in the OPENROUTER_API_KEY variable.
Before first use, install dependencies: cd ~/.claude/skills/openrouter-models/scripts && npm install. All scripts run via npx tsx from the scripts folder. The agent picks the right script from the decision tree: for "what's cheaper" use list-models.ts --sort price, for "who has lower latency" use get-endpoints.ts sorted by latency. The API returns prices in dollars per token — scripts multiply by one million for readability.
What it helps with
- pick the cheapest model in the programming category
- find which provider offers Claude Sonnet with lowest latency
- compare throughput and pricing for GPT-4o, Claude Sonnet 4 and Gemini 2.5 Pro
- list all models that accept images as input
- resolve a vague model name into an exact OpenRouter identifier
Example prompts for the agent
- Show me the five cheapest OpenRouter models
- Which provider has the best uptime for
anthropic/claude-sonnet-4? - Compare pricing and throughput for
openai/gpt-4oandgoogle/gemini-2.5-pro
How to install openrouter-models
1Skills CLI (Vercel)
npx skills add OpenRouterTeam/skills --skill openrouter-models -a claude-code2Manually: put the files in place
~/.claude/skills/openrouter-models/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/openrouter-models/
1Skills CLI (Vercel)
npx skills add OpenRouterTeam/skills --skill openrouter-models -a codex2Manually: put the files in place
.agents/skills/openrouter-models/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add OpenRouterTeam/skills --skill openrouter-models -a cursor2Manually: put the files in place
.cursor/skills/openrouter-models/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install OpenRouterTeam/skills openrouter-models2Manually: put the files in place
.github/skills/openrouter-models/SKILL.md1Skills CLI (Vercel)
npx skills add OpenRouterTeam/skills --skill openrouter-models -a gemini-cli2Platform CLI
gemini skills install https://github.com/OpenRouterTeam/skills --path skills/openrouter-modelsManually: put the files in place
.devin/skills/openrouter-models/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add OpenRouterTeam/skills --skill openrouter-models -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about openrouter-models
How do I install openrouter-models in Claude Code?
Run npx skills add OpenRouterTeam/skills --skill openrouter-models -a claude-code, then navigate to ~/.claude/skills/openrouter-models/scripts and run npm install.
What does openrouter-models do?
The skill queries OpenRouter for the model catalog, pricing, context limits, provider latency and throughput. The agent can search models, compare them, and pick the fastest or cheapest provider.
When is the OpenRouter API key required?
The OPENROUTER_API_KEY variable is only needed for get-endpoints.ts, which shows latency, uptime and throughput by provider. Other scripts work without a key.