DSBench-Hard is DeepSeek's internal hard benchmark for complex coding-agent capabilities. DeepSeek has published model scores but has not disclosed the task count or full methodology.
Browse the latest scores, model modes, release dates, and parameter sizes for DSBench-Hard.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
| Rank | Model | License | |||
|---|---|---|---|---|---|
 DeepSeek-V4-Flash-Vision-Exp Thinking Level · MaxTools | 63.60 | 2026-08-21 | Unknown | — |
DeepSeek reported 63.6 for DeepSeek-V4-Flash-Vision-Exp on August 21, 2026.
