🏆
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 | Qwen3.6-35B-A3B35BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 141,98 tok/s TG Prefill 4.212 · TTFT 12.237 ms | 10× | NVIDIA RTX SPARK (N1X)NVIDIA RTX Spark N1X (6144-core GPU, 20-core CPU) · 128 GB RAM | llama.cppopenclaw_cliQ4_K_M | Details → | |
| 2 | Qwen3.6-35B-A3B35BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 124,89 tok/s TG Prefill 2.476 · TTFT 6.821 ms | 5× | NVIDIA RTX SPARK (N1X)NVIDIA RTX Spark N1X (6144-core GPU, 20-core CPU) · 128 GB RAM | llama.cppopenclaw_cliQ4_K_M | Details → | |
| 3 | Qwen3.6-35B-A3B35BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 67,87 tok/s TG Prefill 1.288 · TTFT 1.499 ms | 1× | NVIDIA RTX SPARK (N1X)NVIDIA RTX Spark N1X (6144-core GPU, 20-core CPU) · 128 GB RAM | llama.cppopenclaw_cliQ4_K_M | Details → | |
| 4 | Qwen3.6-35B-A3B35BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 64,50 tok/s TG Prefill 1.366 · TTFT 1.420 ms | 1× | NVIDIA RTX SPARK (N1X)NVIDIA RTX Spark N1X (6144-core GPU, 20-core CPU) · 128 GB RAM | llama.cppopenclaw_cliQ4_K_M | Details → |