🏆
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-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 364,07 tok/s TG Prefill 2.683 · TTFT 30.535 ms | 10× | NVIDIA RTX PRO 6000 Blackwell Workstation EditionAMD Ryzen 9 9950X 16-Core Processor · 92 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 2 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 302,99 tok/s TG Prefill 2.480 · TTFT 36.311 ms | 10× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 3 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 291,20 tok/s TG Prefill 2.971 · TTFT 34.561 ms | 10× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 4 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 249,27 tok/s TG Prefill 1.839 · TTFT 45.580 ms | 10× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 5 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 197,25 tok/s TG Prefill 5.040 · TTFT 7.497 ms | 10× | 2x NVIDIA RTX A6000AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | vLLMgodclawFP8 | Details → | |
| 6 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 190,16 tok/s TG Prefill 2.465 · TTFT 9.980 ms | 5× | NVIDIA RTX PRO 6000 Blackwell Workstation EditionAMD Ryzen 9 9950X 16-Core Processor · 92 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 7 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 183,48 tok/s TG Prefill 4.576 · TTFT 8.283 ms | 10× | 2x NVIDIA RTX A6000AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | vLLMgodclawFP8 | Details → | |
| 8 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 159,68 tok/s TG Prefill 2.983 · TTFT 10.578 ms | 5× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 9 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 158,21 tok/s TG Prefill 2.237 · TTFT 11.826 ms | 5× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 10 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 154,03 tok/s TG Prefill 1.071 · TTFT 74.122 ms | 10× | 2x NVIDIA RTX A6000AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 11 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 153,61 tok/s TG Prefill 1.063 · TTFT 74.931 ms | 10× | 2x NVIDIA RTX A6000AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 12 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 122,58 tok/s TG Prefill 1.677 · TTFT 15.124 ms | 5× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 13 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 111,56 tok/s TG Prefill 3.912 · TTFT 4.256 ms | 5× | 2x NVIDIA RTX A6000AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | vLLMgodclawFP8 | Details → | |
| 14 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 106,74 tok/s TG Prefill 3.548 · TTFT 4.684 ms | 5× | 2x NVIDIA RTX A6000AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | vLLMgodclawFP8 | Details → | |
| 15 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 79,80 tok/s TG Prefill 936 · TTFT 25.459 ms | 5× | 2x NVIDIA RTX A6000AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 16 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 79,37 tok/s TG Prefill 1.020 · TTFT 24.303 ms | 5× | 2x NVIDIA RTX A6000AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 17 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 71,02 tok/s TG Prefill 597 · TTFT 43.608 ms | 10× | AMD Radeon PRO W7900 Dual SlotAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 18 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 47,77 tok/s TG Prefill 2.291 · TTFT 1.070 ms | 1× | NVIDIA RTX PRO 6000 Blackwell Workstation EditionAMD Ryzen 9 9950X 16-Core Processor · 92 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 19 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 44,78 tok/s TG Prefill 2.079 · TTFT 1.179 ms | 1× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 20 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 39,31 tok/s TG Prefill 1.618 · TTFT 1.513 ms | 1× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 21 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 38,26 tok/s TG Prefill 2.396 · TTFT 1.025 ms | 1× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 22 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 34,68 tok/s TG Prefill 490 · TTFT 25.082 ms | 5× | AMD Radeon PRO W7900 Dual SlotAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 23 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 25,25 tok/s TG Prefill 1.890 · TTFT 1.299 ms | 1× | 2x NVIDIA RTX A6000AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | vLLMgodclawFP8 | Details → | |
| 24 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 25,18 tok/s TG Prefill 261 · TTFT 255.650 ms | 10× | NVIDIA GeForce RTX 5070 TiAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 25 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 24,54 tok/s TG Prefill 1.688 · TTFT 1.452 ms | 1× | 2x NVIDIA RTX A6000AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | vLLMgodclawFP8 | Details → | |
| 26 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 21,84 tok/s TG Prefill 838 · TTFT 2.930 ms | 1× | 2x NVIDIA RTX A6000AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 27 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 21,83 tok/s TG Prefill 830 · TTFT 2.957 ms | 1× | 2x NVIDIA RTX A6000AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 28 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 20,18 tok/s TG Prefill 359 · TTFT 278.984 ms | 10× | NVIDIA GeForce RTX 5090AMD Ryzen 7 5800X3D 8-Core Processor · 126 GB RAM | llama.cppcodex_cliQ4_K_M | Details → | |
| 29 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 19,60 tok/s TG Prefill 427 · TTFT 182.825 ms | 10× | NVIDIA GeForce RTX 3090 TiAMD Ryzen 9 8945HX with Radeon Graphics · 92 GB RAM | llama.cppwebsocketQ4_K_M | Details → | |
| 30 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 19,19 tok/s TG Prefill 463 · TTFT 5.292 ms | 1× | AMD Radeon PRO W7900 Dual SlotAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 31 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 18,48 tok/s TG Prefill 410 · TTFT 83.981 ms | 5× | NVIDIA GeForce RTX 3090 TiAMD Ryzen 9 8945HX with Radeon Graphics · 92 GB RAM | llama.cppwebsocketQ4_K_M | Details → | |
| 32 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 14,12 tok/s TG Prefill 239 · TTFT 121.204 ms | 5× | NVIDIA GeForce RTX 5070 TiAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 33 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 11,93 tok/s TG Prefill 341 · TTFT 124.817 ms | 5× | NVIDIA GeForce RTX 5090AMD Ryzen 7 5800X3D 8-Core Processor · 126 GB RAM | llama.cppcodex_cliQ4_K_M | Details → | |
| 34 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 11,62 tok/s TG Prefill 22 · TTFT 391.135 ms | 5× | 3x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 35 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 11,44 tok/s TG Prefill 22 · TTFT 390.695 ms | 10× | 3x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 36 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 11,43 tok/s TG Prefill 22 · TTFT 391.026 ms | 5× | 3x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 37 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 11,39 tok/s TG Prefill 22 · TTFT 390.917 ms | 10× | 3x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 38 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 11,18 tok/s TG Prefill 22 · TTFT 390.807 ms | 10× | 3x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 39 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 11,12 tok/s TG Prefill 22 · TTFT 390.191 ms | 5× | 3x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 40 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 4,61 tok/s TG Prefill 384 · TTFT 6.371 ms | 1× | NVIDIA GeForce RTX 3090 TiAMD Ryzen 9 8945HX with Radeon Graphics · 92 GB RAM | llama.cppwebsocketQ4_K_M | Details → | |
| 41 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 4,41 tok/s TG Prefill 229 · TTFT 10.683 ms | 1× | NVIDIA GeForce RTX 5070 TiAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 42 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 3,05 tok/s TG Prefill 25 · TTFT 98.876 ms | 1× | 3x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 43 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 3,04 tok/s TG Prefill 25 · TTFT 98.887 ms | 1× | 3x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 44 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 3,04 tok/s TG Prefill 25 · TTFT 98.873 ms | 1× | 3x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 45 | Llama-3.3-Nemotron-Super-49B-v1.549BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 2,48 tok/s TG Prefill 301 · TTFT 8.129 ms | 1× | NVIDIA GeForce RTX 5090AMD Ryzen 7 5800X3D 8-Core Processor · 126 GB RAM | llama.cppcodex_cliQ4_K_M | Details → |