Skip to content
can-i-run-this-llm

Community contribution

Contribute a benchmark

Help everyone check the estimates against real hardware. Run llama-bench or ollama run --verbose, select what you ran and paste the output. No account required.

See the measured benchmarks →

What did you run?

Match the computer, model, quantization and context to your benchmark.

Your computer

Hardware picker is loading. Please wait.

Run setup · optional

These tags help us understand multi-GPU and speculative decoding results.

Paste the output

Paste the full llama-bench table or ollama run --verbose output. We extract the decode tokens per second.

Protected by a small proof-of-work check — no third-party CAPTCHA, no tracking.

Select your computer and paste an output to enable submission.