CRAG
text
+
+
+
+
About
CRAG (Comprehensive RAG Benchmark) is an evaluation framework designed to assess Retrieval-Augmented Generation systems' performance across diverse information retrieval and generation scenarios. This comprehensive benchmark tests AI models' ability to retrieve relevant information and generate accurate, contextually appropriate responses. CRAG provides thorough evaluation of RAG systems' effectiveness in real-world information synthesis and knowledge integration tasks.
+
+
+
+
Evaluation Stats
Total Models3
Organizations1
Verified Results0
Self-Reported3
+
+
+
+
Benchmark Details
Max Score1
Language
en
+
+
+
+
Performance Overview
Score distribution and top performers
Score Distribution
3 models
Top Score
50.3%
Average Score
45.7%
High Performers (80%+)
0Top Organizations
#1Amazon
3 models
45.7%
+
+
+
+
Leaderboard
3 models ranked by performance on CRAG
License | Links | ||||
---|---|---|---|---|---|
Nov 20, 2024 | Proprietary | 50.3% | |||
Nov 20, 2024 | Proprietary | 43.8% | |||
Nov 20, 2024 | Proprietary | 43.1% |