Our new X account is live! Follow @wizwand_team for updates

Text Quality Meta-evaluation on SummEval & Topical-Chat Combined

69.5Overall Score

DeepSeek-V3

Updated 4d ago

Evaluation Results

Method	Links
DeepSeek-V3 2025.02		69.5
GPT-4o 2025.02		69.4
GPT-4 Turbo 2025.02		68.9
GPT-4o mini 2025.02		68.4
Qwen-2.5-72B 2025.02		67.4
Gemma-2-27B 2025.02		66.9
CompassJudger-32B 2025.02		66.7
Phi-4-14B 2025.02		65.8
Llama-3.1-70B 2025.02		65
GPT-3.5 Turbo 2025.02		62.5
Prometheus-2-8x7B 2025.02		59.7
CRITIQUELLM-6B 2025.02		59.6
Prometheus-2-7B 2025.02		59
Auto-J-13B 2025.02		51.4
Prometheus-13B 2025.02		48.4
Themis-8B 2025.02		41.7