Terms of Service

1. Who provides this service

LLM Index (www.llmindex.io, including its public API — the “Service”) is operated by Simon-Pierre Boucher (“we”, “us”). Contact: contact@spboucher.ai. By using the Service you accept these terms.

2. What the Service is

LLM Index publishes independent, methodology-driven evaluations of large language models: scores, confidence intervals, sub-metrics, model responses, judge verdicts and an efficiency frontier. Scores are statistical estimates produced by the published methodology — they are opinions grounded in a documented measurement process, not statements of objective fact about any model or vendor, and they change as new runs complete and the methodology evolves (semver-versioned with a public changelog).

3. Acceptable use

• Respect the published API rate limits (60 requests/minute anonymous; 600 with an issued key). Do not circumvent limits, scrape aggressively, disrupt, probe or overload the Service.
• Do not attempt to extract non-public material (answer keys, anchor item instantiations, credentials) or to interfere with live benchmark runs.
• Do not misrepresent LLM Index data as your own measurement, and do not present altered figures as ours.

4. Using our data & attribution

You may quote, chart and cite scores and rankings from the Service — including commercially — provided you attribute them to “LLM Index (llmindex.io)” with the index version shown alongside the data. Bulk redistribution of the dataset, or products substantially reproducing the Service, require prior written permission. The platform's source code, design, logo and content are © Simon-Pierre Boucher, all rights reserved.

5. Third-party names

Model and provider names and logos appear solely to identify the evaluated systems (nominative use). They remain trademarks of their respective owners; no affiliation, sponsorship or endorsement is implied — by them of us, or by us of them.

6. No warranty · limitation of liability

The Service is provided “as is” and “as available”, without warranties of any kind, express or implied — including accuracy, completeness, fitness for a particular purpose or uninterrupted availability. Do not base high-stakes decisions solely on an index score. To the maximum extent permitted by law, our total liability arising from the use of the Service is limited to the amount you paid us for it (zero for the free Service). Nothing here excludes liability that cannot legally be excluded.

7. Corrections & disputes about scores

Model vendors and users may dispute a result by writing to contact@spboucher.ai with specifics. Substantiated issues are corrected through a new score run (history is immutable) and acknowledged in the public changelog.

8. Changes, suspension, governing law

We may modify or discontinue the Service or these terms at any time; the date above reflects the current version. We may suspend access that violates these terms. These terms are governed by the laws of the Province of Québec and the federal laws of Canada applicable therein; disputes belong to the courts of Québec, Canada.