Versions

4 versions · Metric: CharXiv Accuracy

Versions of CharXiv
VersionReleasedModelsResults
CharXiv Reasoning — Search and code executionConfiguration: Search and code executionJuly 21, 202633
CharXiv Reasoning — Python enabledConfiguration: With PythonDecember 11, 202588
CharXiv Descriptive and reasoning — Mercor single-shotConfiguration: Mercor single-shot agentJune 26, 20243436
CharXiv Reasoning — No toolsConfiguration: No toolsJune 26, 20241616

Recently reported results: CharXiv Reasoning — Search and code execution

Most recent configuration: CharXiv Reasoning — Search and code execution