List your company

Operating System Control on AgentBench OS

37.6AccuracyINFOTREE
1.40810.80420.229.596Oct 28, 2025Nov 28, 2025Dec 30, 2025Jan 31, 2026Mar 3, 2026Apr 4, 2026May 6, 2026

Evaluation results

MethodLinks
INFOTREE2026.0537.6
OpenHands CodeActAgentModel=Qwen-2.5-32B-Coder-Instruct, Training Data=ADP Data2025.1034.7
Tree-GRPO2026.0533.8
Flat GRPO2026.0531.4
OpenHands CodeActAgentModel=Qwen-2.5-32B-Coder-Instruct, Training Data=–2025.1027.8
OpenHands CodeActAgentModel=Qwen-2.5-7B-Coder-Instruct, Training Data=ADP Data2025.1027.1
AgentLMModel=Llama-2-chat-70B, Training Data=AgentInstruct (Zeng et al., 2023)2025.1021.5
OpenHands CodeActAgentModel=Qwen-2.5-14B-Coder-Instruct, Training Data=ADP Data2025.1020.8
AgentLMModel=Llama-2-chat-13B, Training Data=AgentInstruct (Zeng et al., 2023)2025.1018.1
AgentLMModel=Llama-2-chat-7B, Training Data=AgentInstruct (Zeng et al., 2023)2025.1017.4
AgentLMModel=Llama-2-chat-70B, Training Data=–2025.109
AgentLMModel=Llama-2-chat-13B, Training Data=–2025.109
AgentLMModel=Llama-2-chat-7B, Training Data=–2025.108.3
OpenHands CodeActAgentModel=Qwen-2.5-7B-Coder-Instruct, Training Data=–2025.103.5
OpenHands CodeActAgentModel=Qwen-2.5-14B-Coder-Instruct, Training Data=–2025.102.8