🏆
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 | Muse-Glimmer-30B30BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 199,09 tok/s TG Prefill 2.520 · TTFT 2.032 ms | 10× | 4x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppopenclaw_cli batched-bench ROCmQ4_K_M | Details → | |
| 2 | Muse-Glimmer-30B30BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 159,55 tok/s TG Prefill 3.112 · TTFT 1.645 ms | 10× | 4x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppopenclaw_cli batched-bench ROCmQ8_0 | Details → | |
| 3 | Muse-Glimmer-30B30BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 84,05 tok/s TG Prefill 2.716 · TTFT 943 ms | 5× | 4x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppopenclaw_cli batched-bench ROCmQ8_0 | Details → | |
| 4 | Muse-Glimmer-30B30BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 82,79 tok/s TG Prefill 2.200 · TTFT 1.164 ms | 5× | 4x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppopenclaw_cli batched-bench ROCmQ4_K_M | Details → | |
| 5 | Muse-Glimmer-30B30BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 75,28 tok/s TG Prefill 2.774 · TTFT 8.886 ms | 10× | 4x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | vLLMgodclawBF16 | Details → | |
| 6 | Muse-Glimmer-30B30BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 41,70 tok/s TG Prefill 2.092 · TTFT 5.044 ms | 5× | 4x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | vLLMgodclawBF16 | Details → | |
| 7 | Muse-Glimmer-30B30BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 28,00 tok/s TG Prefill 1.219 · TTFT 420 ms | 1× | 4x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppopenclaw_cli batched-bench ROCmQ4_K_M | Details → | |
| 8 | Muse-Glimmer-30B30BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 17,92 tok/s TG Prefill 1.520 · TTFT 337 ms | 1× | 4x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppopenclaw_cli batched-bench ROCmQ8_0 | Details → | |
| 9 | Muse-Glimmer-30B30BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 8,16 tok/s TG Prefill 1.213 · TTFT 1.622 ms | 1× | 4x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | vLLMgodclawBF16 | Details → |