general benchmark
DS-FIM-Eval
DeepSeek's internal Fill-in-the-Middle evaluation dataset for measuring code completion performance improvements in data science contexts
Updated Aug 11, 2026
Models1
Model coverage1
MetricScore
EvidenceB
DS-FIM-Eval Ranking
Higher score ranks better on this benchmark.
| 1 | DE | 78.3% | 100.0% | 1 | C |
DS-FIM-Eval Highlights
The leading models and scores on this benchmark.
What is DS-FIM-Eval?
What DS-FIM-Eval measures and how its scores work.
DeepSeek's internal Fill-in-the-Middle evaluation dataset for measuring code completion performance improvements in data science contexts
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- DS-FIM-Eval
- Modality
- text
- Primary category
- general
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- ds-fim-eval|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about DS-FIM-Eval.
Which model scores highest on DS-FIM-Eval?
DeepSeek-V2.5 is currently ranked first with 78.3%.
What does DS-FIM-Eval measure?
DeepSeek's internal Fill-in-the-Middle evaluation dataset for measuring code completion performance improvements in data science contexts
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.