🏆
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 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 564,60 tok/s TG Prefill 3.577 · TTFT 19.963 ms | 10× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 2 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 561,09 tok/s TG Prefill 2.997 · TTFT 20.809 ms | 10× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 3 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 557,10 tok/s TG Prefill 3.042 · TTFT 20.906 ms | 10× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 4 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 549,60 tok/s TG Prefill 3.637 · TTFT 19.950 ms | 10× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 5 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 308,96 tok/s TG Prefill 3.279 · TTFT 6.536 ms | 5× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 6 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 308,32 tok/s TG Prefill 3.271 · TTFT 6.567 ms | 5× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 7 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 307,99 tok/s TG Prefill 2.795 · TTFT 7.017 ms | 5× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 8 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 307,65 tok/s TG Prefill 2.768 · TTFT 7.045 ms | 5× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 9 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 168,57 tok/s TG Prefill 13.855 · TTFT 2.538 ms | 10× | 4x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | vLLMgodclawGPTQ | Details → | |
| 10 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 104,54 tok/s TG Prefill 2.245 · TTFT 862 ms | 1× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 11 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 104,28 tok/s TG Prefill 2.209 · TTFT 883 ms | 1× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 12 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 103,91 tok/s TG Prefill 2.025 · TTFT 954 ms | 1× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 13 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 103,61 tok/s TG Prefill 2.040 · TTFT 951 ms | 1× | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 14 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 100,64 tok/s TG Prefill 10.369 · TTFT 1.153 ms | 5× | 4x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | vLLMgodclawGPTQ | Details → | |
| 15 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 96,49 tok/s TG Prefill 1.122 · TTFT 20.322 ms | 10× | 2x AMD Radeon PRO W7900 Dual SlotAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 16 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 94,05 tok/s TG Prefill 1.007 · TTFT 9.929 ms | 5× | 2x AMD Radeon PRO W7900 Dual SlotAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 17 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 47,68 tok/s TG Prefill 759 · TTFT 2.540 ms | 1× | 2x AMD Radeon PRO W7900 Dual SlotAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 18 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 40,12 tok/s TG Prefill 233 · TTFT 94.956 ms | 10× | AMD Radeon PRO W7900 Dual SlotAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 19 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 33,94 tok/s TG Prefill 176 · TTFT 56.385 ms | 5× | AMD Radeon PRO W7900 Dual SlotAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 20 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 31,14 tok/s TG Prefill 118 · TTFT 246.235 ms | 10× | NVIDIA GeForce RTX 5070 TiAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 21 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 30,60 tok/s TG Prefill 119 · TTFT 251.486 ms | 10× | NVIDIA GeForce RTX 5070 TiAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 22 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 29,08 tok/s TG Prefill 81 · TTFT 63.546 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 → | |
| 23 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 29,07 tok/s TG Prefill 3.539 · TTFT 554 ms | 1× | 4x AMD Radeon AI PRO R9700AMD Ryzen Threadripper PRO 7955WX 16-Cores · 184 GB RAM | vLLMgodclawGPTQ | Details → | |
| 24 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 26,55 tok/s TG Prefill 139 · TTFT 158.212 ms | 10× | 2x AMD Radeon PRO W7900 Dual SlotAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 25 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 24,87 tok/s TG Prefill 138 · TTFT 262.075 ms | 10× | NVIDIA GeForce RTX 3090 TiAMD Ryzen 9 8945HX with Radeon Graphics · 92 GB RAM | llama.cppwebsocketQ4_K_M | Details → | |
| 26 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 24,12 tok/s TG Prefill 74 · TTFT 69.559 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 → | |
| 27 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 23,94 tok/s TG Prefill 79 · TTFT 32.205 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 → | |
| 28 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 23,62 tok/s TG Prefill 80 · TTFT 331.679 ms | 10× | NVIDIA GeForce RTX 5070 TiAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 29 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 23,44 tok/s TG Prefill 103 · TTFT 96.074 ms | 5× | 2x AMD Radeon PRO W7900 Dual SlotAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 30 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 23,19 tok/s TG Prefill 102 · TTFT 136.966 ms | 5× | NVIDIA GeForce RTX 5070 TiAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 31 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 22,55 tok/s TG Prefill 103 · TTFT 138.591 ms | 5× | NVIDIA GeForce RTX 5070 TiAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 32 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 20,05 tok/s TG Prefill 68 · TTFT 189.332 ms | 5× | NVIDIA GeForce RTX 5070 TiAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 33 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 19,35 tok/s TG Prefill 148 · TTFT 13.078 ms | 1× | AMD Radeon PRO W7900 Dual SlotAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 34 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 18,77 tok/s TG Prefill 126 · TTFT 135.104 ms | 5× | NVIDIA GeForce RTX 3090 TiAMD Ryzen 9 8945HX with Radeon Graphics · 92 GB RAM | llama.cppwebsocketQ4_K_M | Details → | |
| 35 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 17,52 tok/s TG Prefill 73 · TTFT 35.131 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 → | |
| 36 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 14,89 tok/s TG Prefill 89 · TTFT 21.613 ms | 1× | 2x AMD Radeon PRO W7900 Dual SlotAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 37 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 12,11 tok/s TG Prefill 75 · TTFT 158.430 ms | 5× | NVIDIA RTX PRO 6000 Blackwell Workstation EditionAMD Ryzen 9 9950X 16-Core Processor · 92 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 38 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 10,46 tok/s TG Prefill 67 · TTFT 7.648 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 → | |
| 39 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 9,28 tok/s TG Prefill 69 · TTFT 119.823 ms | 5× | NVIDIA RTX PRO 6000 Blackwell Workstation EditionAMD Ryzen 9 9950X 16-Core Processor · 92 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 40 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 9,23 tok/s TG Prefill 75 · TTFT 110.236 ms | 10× | NVIDIA RTX PRO 6000 Blackwell Workstation EditionAMD Ryzen 9 9950X 16-Core Processor · 92 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 41 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 8,99 tok/s TG Prefill 74 · TTFT 109.505 ms | 10× | NVIDIA RTX PRO 6000 Blackwell Workstation EditionAMD Ryzen 9 9950X 16-Core Processor · 92 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 42 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 8,29 tok/s TG Prefill 86 · TTFT 99.324 ms | 10× | NVIDIA GeForce RTX 5090AMD Ryzen 7 5800X3D 8-Core Processor · 126 GB RAM | llama.cppcodex_cliQ4_K_M | Details → | |
| 43 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 8,29 tok/s TG Prefill 86 · TTFT 99.977 ms | 10× | NVIDIA GeForce RTX 5090AMD Ryzen 7 5800X3D 8-Core Processor · 126 GB RAM | llama.cppcodex_cliQ4_K_M | Details → | |
| 44 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 8,20 tok/s TG Prefill 98 · TTFT 19.630 ms | 1× | NVIDIA GeForce RTX 5070 TiAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ4_K_M | Details → | |
| 45 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 8,18 tok/s TG Prefill 85 · TTFT 100.249 ms | 5× | NVIDIA GeForce RTX 5090AMD Ryzen 7 5800X3D 8-Core Processor · 126 GB RAM | llama.cppcodex_cliQ4_K_M | Details → | |
| 46 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 8,17 tok/s TG Prefill 86 · TTFT 99.836 ms | 5× | NVIDIA GeForce RTX 5090AMD Ryzen 7 5800X3D 8-Core Processor · 126 GB RAM | llama.cppcodex_cliQ4_K_M | Details → | |
| 47 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 7,93 tok/s TG Prefill 63 · TTFT 8.091 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 → | |
| 48 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 7,24 tok/s TG Prefill 64 · TTFT 30.093 ms | 1× | NVIDIA GeForce RTX 5070 TiAMD Ryzen Threadripper PRO 5975WX 32-Cores · 247 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 49 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 6,96 tok/s TG Prefill 123 · TTFT 15.650 ms | 1× | NVIDIA GeForce RTX 3090 TiAMD Ryzen 9 8945HX with Radeon Graphics · 92 GB RAM | llama.cppwebsocketQ4_K_M | Details → | |
| 50 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 6,29 tok/s TG Prefill 52 · TTFT 164.098 ms | 5× | NVIDIA GeForce RTX 5090AMD Ryzen 7 5800X3D 8-Core Processor · 126 GB RAM | llama.cppcodex_cliQ8_0 | Details → | |
| 51 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 6,22 tok/s TG Prefill 52 · TTFT 164.464 ms | 10× | NVIDIA GeForce RTX 5090AMD Ryzen 7 5800X3D 8-Core Processor · 126 GB RAM | llama.cppcodex_cliQ8_0 | Details → | |
| 52 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 4,93 tok/s TG Prefill 64 · TTFT 30.298 ms | 1× | NVIDIA RTX PRO 6000 Blackwell Workstation EditionAMD Ryzen 9 9950X 16-Core Processor · 92 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 53 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 4,09 tok/s TG Prefill 77 · TTFT 24.992 ms | 1× | NVIDIA GeForce RTX 5090AMD Ryzen 7 5800X3D 8-Core Processor · 126 GB RAM | llama.cppcodex_cliQ4_K_M | Details → | |
| 54 | Qwen3.5-122B-A10B122BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 3,38 tok/s TG Prefill 48 · TTFT 40.581 ms | 1× | NVIDIA GeForce RTX 5090AMD Ryzen 7 5800X3D 8-Core Processor · 126 GB RAM | llama.cppcodex_cliQ8_0 | Details → |