Search, inspect, and benchmark 300+ AI models from your editor.
Add to .cursor/mcp.json or .vscode/mcp.json:
{
"mcpServers": {
"index9": {
"command": "npx",
"args": ["-y", "@index9/mcp"]
}
}
}claude mcp add --transport stdio index9 -- npx -y @index9/mcp| Tool | Description | API Key |
|---|---|---|
find_models |
Search models by query, price, context window, or capabilities | No |
get_models |
Get pricing, limits, and capabilities for any model by ID or alias | No |
test_model |
Run your prompt across multiple models — compare outputs, latency, and cost | Optional |
test_model with dryRun=true estimates cost without inference (no key needed). Live tests require an OpenRouter API key:
{
"mcpServers": {
"index9": {
"command": "npx",
"args": ["-y", "@index9/mcp"],
"env": {
"OPENROUTER_API_KEY": "sk-..."
}
}
}
}Your key is passed per-request and never stored.
- index9.dev — full docs and examples
- Issues