List the current final ranking, look up one model across available surfaces, ask how its rank changed, or retrieve a saved ranking at a date. The same API key works with REST and MCP.
Create your free Data accountWhat you can query
| Data | Fields and coverage | Plan | MCP tool |
|---|---|---|---|
| Account usage | Current plan, successful and remaining monthly reads, request-rate usage, and both reset times. Usage checks do not consume a monthly read. | Free | benchlm_get_usage |
| Current BenchAlign rankings | Final scores, published ranks, evidence labels, interval bounds, scoring date, and methodology version for each available surface. | Free | benchlm_list_current_rankings |
| Current model rankings | One model’s final score and rank across every available BenchAlign surface, with explicit gaps where it is unranked. | Free | benchlm_get_current_model_rankings |
| Model catalog | Names, creators, canonical IDs, source links, and sourced release dates with their date precision. | Free | benchlm_list_models / benchlm_get_model |
| Benchmark catalog | Names, categories, catalog IDs, and source references. | Free | benchlm_list_benchmarks |
| Historical coverage | Available methods, retained date windows, and delivery coverage. Check this before requesting history. | Free | benchlm_history_coverage |
| BenchAlign history | Saved BenchLM scores, ranks, evidence labels, ranking surfaces, and methodology versions for an exact model. | Data Pro | benchlm_rank_history |
| Ranking at a date | Stored BenchAlign ranking rows as of a UTC date. The response identifies incomplete coverage and preserves the saved scoring version. | Data Pro | benchlm_ranking_snapshot |
| Benchmark history | Cleared measurements for an exact model, benchmark, and protocol. Availability differs by source; this is not every score on the website. | Data Pro | benchlm_benchmark_history |
| Benchmark changes | First recorded catalog definitions and results, additions, removals, and reintroductions where retained. | Data Pro | benchlm_benchmark_events |
Current ranking responses contain final BenchAlign outputs, not the underlying benchmark or upstream ranking scores. Historical dates describe saved repository states, not when a publisher first released a result. Missing records stay missing. Individual benchmark history contains measurements, not historical leaderboard positions.
Plans and allowances
- Free · $0
- 1,000 successful reads per account per month. Current BenchAlign rankings, model and benchmark catalogs, coverage checks, and up to 10 requests per minute. No payment details required.
- Data Pro · $49 USD/month
- 100,000 successful reads per month, the same current rankings and catalogs, historical queries within available coverage, and up to 60 requests per minute. Check coverage in your account before subscribing.
- Radar + Data · $59 USD/month
- Radar and Data Pro billed together, with separate API keys and allowances. Link your account from Radar. Existing monthly, annual, and legacy subscriptions retain their terms until period end before bundle checkout opens.
Free allowances renew on the account's activation anniversary each month. Paid allowances follow the subscription period. Short months use their last day. All keys for one account share usage across REST and MCP; rotating keys does not reset it.
One successful page uses one read. Initialization, discovery, coverage and usage checks, and unsuccessful queries do not use the monthly allowance. Check usage with GET /v1/usage or benchlm_get_usage; both report monthly and request-rate reset times. Historical queries return up to 500 rows per page, with up to 90 days per request and a cursor for the next page.
Connect a client
Sign in to your Data account and create an API key. In your MCP client, choose Streamable HTTP, use https://data.benchlm.ai/mcp, and send Authorization: Bearer YOUR_DATA_KEY. Keep the key in the client's secret settings.
curl 'https://data.benchlm.ai/v1/rankings/current?surface=overall&limit=10' \
-H 'Authorization: Bearer YOUR_DATA_KEY'curl 'https://data.benchlm.ai/mcp' \
-H 'Authorization: Bearer YOUR_DATA_KEY' \
-H 'Content-Type: application/json' \
-H 'Accept: application/json, text/event-stream' \
-H 'MCP-Protocol-Version: 2025-11-25' \
--data '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"benchlm_list_current_rankings","arguments":{"surface":"overall","limit":10}}}'Current ranking pages take surface, limit, and offset. Resolve exact IDs with the catalog tools before requesting history. Benchmark history takes modelKey, category, and benchmarkKey. Rank history takes modelKey; a ranking snapshot takes at.
The account query reference lists each REST endpoint and MCP tool. Current and catalog pages contain at most 100 records. Historical responses include dates, provenance, gaps, and applicable source notices. Supported MCP protocol versions are 2026-07-28, 2025-11-25, and 2025-06-18.
Data is a separate product from Radar. For the website's existing downloads and their terms, see the dataset page.