agents benchmark
DSBench-FullStack
DSBench-FullStack is DeepSeek's internal full-stack development test set for evaluating coding agents on end-to-end software engineering tasks.
Updated Aug 11, 2026
Models1
Model coverage1
MetricScore
EvidenceB
DSBench-FullStack Ranking
Higher score ranks better on this benchmark.
| 01 | DE | 68.7% | 100.0% | 1 | C |
DSBench-FullStack Highlights
The leading models and scores on this benchmark.
What is DSBench-FullStack?
What DSBench-FullStack measures and how its scores work.
DSBench-FullStack is DeepSeek's internal full-stack development test set for evaluating coding agents on end-to-end software engineering tasks.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- DSBench-FullStack
- Modality
- text
- Primary category
- agents
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- dsbench-fullstack|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about DSBench-FullStack.
Which model scores highest on DSBench-FullStack?
DeepSeek-V4-Flash-0731 is currently ranked first with 68.7%.
What does DSBench-FullStack measure?
DSBench-FullStack is DeepSeek's internal full-stack development test set for evaluating coding agents on end-to-end software engineering tasks.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 model results are currently shown.
Does this benchmark affect the overall score?
No. This benchmark is shown for reference but does not contribute to the overall score.