List your company

Proactive Agent Task Execution on Pare-Bench 1.0 (full)

64.3Success@4Gemini 3 Flash
5.43620.7183651.282Apr 1, 2026

Evaluation results

MethodLinks
Gemini 3 Flash2026.0464.316.142.119.167.121.7
Claude 4.5 Sonnet2026.0460.818.24212.878.220.2
Gemini 3 Pro2026.0459.411.935.116.571.219.5
GPT-52026.0457.317.537.428.170.220.6
Qwen 3 4B Instruct2026.04356.318.520.563.716.7
Llama 3.2 3B Instruct2026.0423.81.4102358.416.6
Gemma 3 4B Instruct2026.047.70.7314.217.68.8