List your company

Proactive Assistance on Pro2Bench Average

83G-Mean F1Qwen3.6-VL-27B
3.9624.484565.52Jun 3, 2026

Evaluation results

MethodLinks
Qwen3.6-VL-27BFT=true, PWR=true, Architecture=split duplex+planner2026.068347
Llama 4-MaverickFT=true, PWR=true, Architecture=split duplex+planner2026.067663
Llama 4-MaverickFT=true, Architecture=monolithic2026.066747
Claude Opus 4.6Architecture=monolithic, Mode=zero-shot2026.065132
Gemini 3.1 ProArchitecture=monolithic, Mode=zero-shot2026.064835
Qwen3.6-VL-27BFT=true, Architecture=monolithic2026.064535
Llama 4-MaverickArchitecture=monolithic2026.064535
Qwen3-VL-235BArchitecture=monolithic2026.064330
Qwen3.6-VL-27BArchitecture=monolithic2026.063137
GPT-5.2Architecture=monolithic, Mode=zero-shot2026.067—