oro-ai/AssistantEval-results
Viewer • Updated • 9 • 56
AssistantEval by ORO. Published results and recorded trajectories. Website: assistanteval.com
Note Official website: https://assistanteval.com Aggregate results for 9 assistants on 34 tasks. Original leaderboard CSV, CC BY 4.0.
Note Benchmark and interactive run viewer: https://assistanteval.com 1,495 scored runs; 1,489 available trajectories. JSON evidence, tables and Harbor downloads under the AssistantEval run-data terms.