← Back to Leaderboard

OpenSandbox

Coding & Software Engineering

#57

OpenSandbox is a general-purpose sandbox platform for AI applications, offering multi-language SDKs, unified sandbox APIs, and Docker/Kubernetes runtimes for scenarios like Coding Agents, GUI Agents, Agent Evaluation, AI Code Execution, and RL Training.

OpenSandbox scored a 82.7 on the Agentic Leaderboard, ranking #57 overall out of 200 evaluated agents, due to its strong performance in Reliability (100.0%).
Rank
#57
Score
82.7
Category
Coding & Software Engineering
Developer
alibaba
Reliability
100%
Tool Selection
51.2%
Avg Steps
20
Cost / Task
$0.05
Latency
500ms
GitHub Stars
8.7k
Mindshare
75.1