🏆
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 | Nemotron-3.5-Lightning-30B-A3B30BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 146,86 tok/s TG Prefill 2.806 · TTFT 12.191 ms | 10× | NVIDIA RTX SPARK (N1X)NVIDIA RTX Spark N1X (6144-core GPU, 20-core CPU) · 128 GB RAM | llama.cppopenclaw_cliQ4_0 | Details → | |
| 2 | Nemotron-3.5-Lightning-30B-A3B30BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 121,80 tok/s TG Prefill 1.943 · TTFT 7.408 ms | 5× | NVIDIA RTX SPARK (N1X)NVIDIA RTX Spark N1X (6144-core GPU, 20-core CPU) · 128 GB RAM | llama.cppopenclaw_cliQ4_0 | Details → | |
| 3 | Nemotron-3.5-Lightning-30B-A3B30BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 72,08 tok/s TG Prefill 1.293 · TTFT 1.669 ms | 1× | NVIDIA RTX SPARK (N1X)NVIDIA RTX Spark N1X (6144-core GPU, 20-core CPU) · 128 GB RAM | llama.cppopenclaw_cliQ4_0 | Details → | |
| 4 | Nemotron-3.5-Lightning-30B-A3B30BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 66,20 tok/s TG Prefill 1.300 · TTFT 1.660 ms | 1× | NVIDIA RTX SPARK (N1X)NVIDIA RTX Spark N1X (6144-core GPU, 20-core CPU) · 128 GB RAM | llama.cppopenclaw_cliQ4_0 | Details → |