Multi-turn Conversation Evaluation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
81.1MT-Bench Score
68
May 12, 2026
3.82Speedup
25
Mar 4, 2026
5.6MT-Bench Score
20
May 25, 2026
8.3MT-Bench Score
20
May 25, 2026
6.1MT-Bench Score
20
May 25, 2026
95Actions Score
9
May 27, 2026
8.538GPT-4 Score
5
Feb 26, 2026