DeepSeek-V3.1

🧠 Hybrid inference: Think & Non-Think — one model, two modes⚡️ Faster thinking: DeepSeek-V3.1-Think reaches answers in less time vs. DeepSeek-R1-0528🛠️ Stronger agent skills: Post-training boosts tool use and multi-step agent tasks
TypeChat(Thinking)OrganizationDeepSeekRelease Date2025-08-20
Websitehttps://www.deepseek.com/

Ability Scores

Evaluation Case Report

Indicator NameIndicator WeightEvaluation TypePassed CasesFailed CasesPass RateCase Details
Logical Equivalence3hybrid10377%
Optimization Depth4subjective151452%
Syntax Error Detection2hybrid12192%

Evaluation Process