BenchGecko

Measured conclusion: Alive, not deep-tested. L2 outcome: not_attempted.

AI model prices per provider, benchmarks, own AI behavior tests and AI economy data, with sources.

Add to your agent

Copy this into your MCP client config:

{
  "mcpServers": {
    "BenchGecko": {
      "url": "https://benchgecko.ai/api/mcp",
      "transport": "streamable-http"
    }
  }
}

Measured tools (9)

• search_models read-only

  Find AI models in the BenchGecko catalog by name, slug or lab. Returns slugs to use with get_model, cheapest_provider and compare_models, with BenchGecko score and list price.

  Parameters:

  • query (required) — string — Model name, slug or lab, e.g. "claude opus", "gpt-5", "deepseek"
  • limit — integer

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• get_model read-only

  Core facts for one AI model: BenchGecko score and rank, list price, price at every provider, benchmark scores with their original sources, Gecko Tests grades and as-of dates.

  Parameters:

  • model (required) — string — Model slug (from search_models) or name, e.g. "claude-opus-5-5"

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• cheapest_provider read-only

  Every provider serving a model, sorted by input price per 1M tokens (refreshed daily from provider endpoints), with output price, quantization and 1 day uptime.

  Parameters:

  • model (required) — string — Model slug or name

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• compare_models read-only

  Side by side: BenchGecko score, list price, cheapest provider, context window, shared benchmark scores and Gecko Tests grades for two models.

  Parameters:

  • a (required) — string — First model slug or name
  • b (required) — string — Second model slug or name

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• get_gecko_test read-only

  Results of one of BenchGecko's own tests (own measurements, CC BY 4.0): who-are-you (Does the model know which lab made it?) world-map (How well does the model draw the world map from memory?) censorship-index (How often does the model refuse legitimate questions?) knowledge-horizon (Where does the model's knowledge of world events actually stop?) tokenizer-tax (How many more tokens does the same text cost outside English?) same-model-different-host (Do providers serving the same open model give the same quality?) model-drift-index (Do models quietly change behind the same name?)

  Parameters:

  • test (required) — string
  • limit — integer — Rows in the text summary (structured result has all rows)

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• get_scorecard read-only

  BenchGecko Gecko Scorecard: every model across BenchGecko's own profile tests with a letter grade per test and an overall Gecko Score (behavior, not intelligence).

  Parameters:

  • limit — integer

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• latest_findings read-only

  Latest notable results detected in the Gecko Tests and price data (fixed rules, numbers and quotes from stored results).

  Parameters:

  • limit — integer

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• search read-only

  Search BenchGecko models and Gecko Tests. Returns ids for fetch.

  Parameters:

  • query (required) — string

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• fetch read-only

  Fetch the full record for an id returned by search (model:<slug>, gecko-test:<slug> or scorecard).

  Parameters:

  • id (required) — string

  Hints: read-only yes · destructive no · idempotent yes · open-world no

Repository

Measured by MCPYP probing. Untested fields are omitted, never estimated. Listing is always free; MCPYP never proxies calls or stores user tokens.