llm-speed Terms of Use

Last updated 2026-10-22. These terms cover llm-speed.com, the API at api.llm-speed.com (including the assistant lookups used by the llm-speed ChatGPT app, Claude connector and Custom GPT) and the published datasets. The CLI is covered by its own open-source license.

1. What the service is

llm-speed publishes measured speed and memory results for running open-weight language models on local hardware: tokens per second, time to first token, peak memory and whether a model fits on a GPU or Mac at a given quant and context. Results are labelled as measured by llm-speed or community-submitted. llm-speed measures speed and memory, not model quality.

2. No warranty

Benchmark results describe the specific hardware, software build, settings and date they were measured with. Your results can differ. The service and data are provided "as is", without warranties of any kind, express or implied, including accuracy, fitness for a particular purpose and availability. Don't rely on them as the only basis for a purchase.

3. Assistant answers

When you use llm-speed through ChatGPT, Claude or another assistant, the assistant writes the answer; our tools only return data. Check the linked llm-speed page for the measured result and its conditions. The assistant provider's own terms and privacy policy apply to your conversation.

4. Acceptable use

Don't use the service to break the law, to attack or overload it (the API is rate-limited per client), to submit fabricated or test results as real measurements, or to impersonate llm-speed. We may block traffic or remove submissions that break these rules.

5. Data license

Benchmark data is licensed under CC BY 4.0 (see llm-speed.com/about). Attribute it to "llm-speed (llm-speed.com)" and link the page you used where practical.

6. Liability

To the extent the law allows, llm-speed is not liable for indirect or consequential losses, or for decisions made from the data. Nothing here limits liability that can't be limited by law.

7. Changes and contact

We may update these terms; the date above changes when we do. Questions or problems: open an issue on the public CLI repository (github.com/meadow-kun/llm-speed). Privacy questions: see llm-speed.com/privacy.