Created bymario-alka.dePowered bynoob2claw.detricoma.de
🏆
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
Reset
Metric:
#Model / MakerMetricsParallelGPU / CPU / RAMRuntime
1Qwen3-30B-A3B-Thinking-250730BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
161,05 tok/s TG
Prefill 5.256 · TTFT 6.170 ms
10×NVIDIA DGX SPARK (GB10) iNVIDIA Grace · 120 GB RAMllama.cppgodclawQ4_K_M
2Qwen3-30B-A3B-Thinking-250730BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
129,46 tok/s TG
Prefill 3.442 · TTFT 3.808 ms
5×NVIDIA DGX SPARK (GB10) iNVIDIA Grace · 120 GB RAMllama.cppgodclawQ4_K_M
3Qwen3-30B-A3B-Thinking-250730BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
81,15 tok/s TG
Prefill 2.472 · TTFT 986 ms
1×NVIDIA DGX SPARK (GB10) iNVIDIA Grace · 120 GB RAMllama.cppgodclawQ4_K_M