🤖
Harness benchmark
How well does the model solve real tasks?
Quality on real agent and chat tasks – the model must handle real tasks with tools and multiple steps. Scored by success rate and task points achieved.
🎯 Success rate🏆 Task points🔧 Tool use🧠 real tasks
| # | Model / Maker | Metrics | GPU / CPU / RAM | Runtime | ||
|---|---|---|---|---|---|---|
| 1 | Gemma-4-26B-A4B-it26BGoogle Harness benchmarkTool Usage Standard 1.0 | 2.701 Pkt 90,6% · 16/16 Aufg. · 118,0 s | NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen 9 9950X 16-Core Processor · 92 GB RAM | openclaw_cli | Details → | |
| 2 | Devstral-Small-2-24B-Instruct-251224BMistral AI Harness benchmarkTool Usage Standard 1.0 | 2.597 Pkt 87,1% · 16/16 Aufg. · 165,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 31 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 3 | Gemma-4-26B-A4B-it26BGoogle Harness benchmarkTool Usage Standard 1.0 | 2.440 Pkt 81,9% · 16/16 Aufg. · 218,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 123 GB RAM | openclaw_cli | Details → | |
| 4 | MiniMax-M2.5230BMiniMax Harness benchmarkTool Usage Standard 1.0 | 2.427 Pkt 81,4% · 16/16 Aufg. · 228,0 s | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | vLLMgodclawBF16 | Details → | |
| 5 | Gemma-4-26B-A4B-it26BGoogle Harness benchmarkTool Usage Standard 1.0 | 2.383 Pkt 80,0% · 16/16 Aufg. · 245,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 123 GB RAM | openclaw_cli | Details → | |
| 6 | MiniMax-M2.5230BMiniMax Harness benchmarkTool Usage Standard 1.0 | 2.340 Pkt 78,5% · 16/16 Aufg. · 266,0 s | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | openclaw_cli | Details → | |
| 7 | Gemma-4-26B-A4B-it26BGoogle Harness benchmarkTool Usage Standard 1.0 | 2.268 Pkt 76,1% · 16/16 Aufg. · 347,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 31 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 8 | Devstral-Small-2-24B-Instruct-251224BMistral AI Harness benchmarkGodBrain MCP Deep 40 - Wissensbasis-Lebenszyklus 1.0 | 5.738 Pkt 75,7% · 40/40 Aufg. · 1.879,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 31 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 9 | Gemma-4-26B-A4B-it26BGoogle Harness benchmarkTool Usage Standard 1.0 | 2.197 Pkt 73,7% · 16/16 Aufg. · 375,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 31 GB RAM | openclaw_cli | Details → | |
| 10 | MiniMax-M2.5230BMiniMax Harness benchmarkTool Usage V 1.0 | 1.245 Pkt 70,7% · 15/15 Aufg. · 301,0 s | NVIDIA GB10 (DGX Spark)NVIDIA Grace · 120 GB RAM | openclaw_cli | Details → | |
| 11 | Devstral-Small-2-24B-Instruct-251224BMistral AI Harness benchmarkTool Usage Extrem 40 - Tech-Radar Mission 1.0 | 5.629 Pkt 68,6% · 40/40 Aufg. · 2.018,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 31 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 12 | MiniMax-M2.5230BMiniMax Harness benchmarkTool Usage Standard 1.0 | 2.033 Pkt 68,2% · 16/16 Aufg. · 373,0 s | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | vLLMgodclawBF16 | Details → | |
| 13 | MiniMax-M2.5230BMiniMax Harness benchmarkTools Usage Mini 1.0 | 130 Pkt 65,0% · 2/2 Aufg. · 70,0 s | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | godclaw | Details → | |
| 14 | MiniMax-M2.5230BMiniMax Harness benchmarkTool Usage V 1.0 | 1.079 Pkt 61,3% · 15/15 Aufg. · 457,0 s | 2x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | openclaw_cli | Details → | |
| 15 | MiniMax-M2.7230BMiniMax Harness benchmarkTool-Parcours Dossier 40 V1.0 | 3.641 Pkt 61,1% · 40/40 Aufg. · 8.209,0 s | NVIDIA GB10 (DGX Spark)NVIDIA Grace · 120 GB RAM | vLLMgodclawAWQ | Details → | |
| 16 | MiniMax-M2.5230BMiniMax Harness benchmarkTool-Parcours Dossier 40 V1.0 | 3.622 Pkt 60,8% · 34/40 Aufg. · 7.862,0 s | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | vLLMgodclawBF16 | Details → | |
| 17 | MiniMax-M2.5230BMiniMax Harness benchmarkTool Usage Standard 1.0 | 1.762 Pkt 59,1% · 16/16 Aufg. · 526,0 s | NVIDIA GB10 (DGX Spark)NVIDIA Grace · 120 GB RAM | openclaw_cli | Details → | |
| 18 | MiniMax-M2.5230BMiniMax Harness benchmarkTool Usage Standard 1.0 | 1.714 Pkt 57,5% · 16/16 Aufg. · 487,0 s | 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | vLLMgodclawBF16 | Details → | |
| 19 | MiniMax-M2.5230BMiniMax Harness benchmarkTools Usage Mini 1.0 | 106 Pkt 53,0% · 2/2 Aufg. · 71,0 s | 2x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | openclaw_cli | Details → | |
| 20 | MiniMax-M2.5230BMiniMax Harness benchmarkTextgenerierung V 1.0 | 790 Pkt 52,0% · 15/15 Aufg. · 261,0 s | NVIDIA GB10 (DGX Spark)NVIDIA Grace · 120 GB RAM | openclaw_cli | Details → | |
| 21 | Qwen3.6-27B27BQwen (Alibaba) Harness benchmarkTool Usage Standard 1.0 | 1.220 Pkt 40,9% · 16/16 Aufg. · 1.036,0 s | NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen 9 9950X 16-Core Processor · 92 GB RAM | vLLMgodclaw | Details → | |
| 22 | Qwen3.6-27B27BQwen (Alibaba) Harness benchmarkTool-Parcours Dossier 40 V1.0 | 2.267 Pkt 38,0% · 38/40 Aufg. · 10.331,0 s | NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen 9 9950X 16-Core Processor · 92 GB RAM | vLLMgodclaw | Details → | |
| 23 | Gemma-4-26B-A4B-it26BGoogle Harness benchmarkGodBrain MCP Deep 40 - Wissensbasis-Lebenszyklus 1.0 | 2.271 Pkt 30,0% · 18/40 Aufg. · 982,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 31 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 24 | MiniMax-M2.7230BMiniMax Harness benchmarkTool Usage Standard 1.0 | 858 Pkt 28,8% · 16/16 Aufg. · 1.222,0 s | NVIDIA GB10 (DGX Spark)NVIDIA Grace · 120 GB RAM | godclaw | Details → | |
| 25 | Gemma-4-26B-A4B-it26BGoogle Harness benchmarkTool-Parcours Dossier 40 V1.0 | 1.489 Pkt 25,0% · 22/40 Aufg. · 8.783,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 31 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 26 | Devstral-Small-2-24B-Instruct-251224BMistral AI Harness benchmarkTool-Parcours Dossier 40 V1.0 | 1.436 Pkt 24,1% · 21/40 Aufg. · 8.887,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 31 GB RAM | llama.cppgodclawQ8_0 | Details → | |
| 27 | Qwen2.5-32B-Instruct-AWQ32BQwen (Alibaba) Harness benchmarkTool Usage V 1.0 | 235 Pkt 13,4% · 15/15 Aufg. · 1.783,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 31 GB RAM | openclaw_cli | Details → | |
| 28 | gemma-4-31B-it31BGoogle Harness benchmarkTool Usage Standard 1.0 | 298 Pkt 10,0% · 16/16 Aufg. · 9.625,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 123 GB RAM | openclaw_cli | Details → | |
| 29 | Qwen2.5-32B-Instruct-AWQ32BQwen (Alibaba) Harness benchmarkTool Usage Standard 1.0 | 298 Pkt 10,0% · 16/16 Aufg. · 4.687,0 s | AMD Radeon 8060S GraphicsAMD RYZEN AI MAX+ 395 w/ Radeon 8060S · 123 GB RAM | openclaw_cli | Details → | |
| 30 | MiniMax-M2.5230BMiniMax Harness benchmarkTools Usage Mini 1.0 | 10 Pkt 10,0% · 1/1 Aufg. · 65,0 s | 2x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionAMD Ryzen Threadripper PRO 9965WX 24-Cores · 125 GB RAM | openclaw_cli | Details → |