🏆
Rankings
LLM Leaderboard
Real measurements of every tested language model on real, named hardware – rated by raw speed (performance in tokens per second, prefill and time to first token) and by practical task quality in complete agent and chat runs (harness). Pick a benchmark type below or filter by model, maker and hardware to see exactly what is tested and how the results are produced.
⚡ Performance (tok/s)🤖 Harness quality👥 Concurrency🖥️ real hardware
| # | Model / Maker | Metrics | Parallel | GPU / CPU / RAM | Runtime | ||
|---|---|---|---|---|---|---|---|
| 1 | Llama-3.2-3B-Instruct3BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 226,59 tok/s TG Prefill 9.526 · TTFT 3.488 ms | 10× | 2x NVIDIA GeForce RTX 20604x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz · 504 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 2 | Llama-3.2-3B-Instruct3BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 215,97 tok/s TG Prefill 9.775 · TTFT 3.283 ms | 10× | 2x NVIDIA GeForce RTX 20604x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz · 504 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 3 | Llama-3.2-3B-Instruct3BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 173,08 tok/s TG Prefill 7.158 · TTFT 2.024 ms | 5× | 2x NVIDIA GeForce RTX 20604x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz · 504 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 4 | Llama-3.2-3B-Instruct3BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 158,90 tok/s TG Prefill 7.616 · TTFT 1.908 ms | 5× | 2x NVIDIA GeForce RTX 20604x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz · 504 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 5 | Llama-3.2-3B-Instruct3BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 100,62 tok/s TG Prefill 3.646 · TTFT 677 ms | 1× | 2x NVIDIA GeForce RTX 20604x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz · 504 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 6 | Llama-3.2-3B-Instruct3BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 70,69 tok/s TG Prefill 3.840 · TTFT 645 ms | 1× | 2x NVIDIA GeForce RTX 20604x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz · 504 GB RAM | llama.cppgodclawQ8_0 | Details → |