LegalBench ↗77.73 %
DeepSeek V3 (03/24/2025) · LegalBench · Accuracy ↗
DeepSeek / MODEL EVIDENCE
DeepSeek V3 (03/24/2025)
10 benchmarks · 10 observations · 10 sources
Scores at a glance
10 benchmarks · large dot: DeepSeek V3 (03/24/2025); small dots: closest models where reported
CorpFin v2 ↗54.74 %
DeepSeek V3 (03/24/2025) · CorpFin · Accuracy ↗
DeepSeek V3 (03/24/2025) · TaxEval · Accuracy ↗
MedQA (USMLE) ↗82 %
DeepSeek V3 (03/24/2025) · MedQA · Accuracy ↗
AIME 2024–2025 · Vals ↗52.2 %
DeepSeek V3 (03/24/2025) · AIME · Accuracy ↗
MATH-500 ↗88.6 %
DeepSeek V3 (03/24/2025) · MATH 500 · Accuracy ↗
MGSM (Multilingual GSM) ↗91.67 %
DeepSeek V3 (03/24/2025) · MGSM · Accuracy ↗
GPQA Diamond ↗61.62 %
DeepSeek V3 (03/24/2025) · GPQA Diamond · Accuracy ↗
MMLU-Pro ↗79.47 %
DeepSeek V3 (03/24/2025) · MMLU Pro · Accuracy ↗
LiveCodeBench ↗65.48 %
DeepSeek V3 (03/24/2025) · LiveCodeBench · Accuracy ↗
Closest models by score
Similar positions on shared benchmark leaderboards. Smaller gaps are closer.
GPT-4.1 Mini →OpenAI
10 of 10 shared benchmark comparisons
9 of 10 shared benchmark comparisons
10 of 10 shared benchmark comparisons
How similarity is measured
Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.
Sources
Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.
- AIME — evaluation methodology and results ↗
- CorpFin — evaluation methodology and results ↗
- GPQA Diamond — evaluation methodology and results ↗
- LiveCodeBench — evaluation methodology and results ↗
- LegalBench — evaluation methodology and results ↗
- MATH 500 — evaluation methodology and results ↗
- MedQA — evaluation methodology and results ↗
- MGSM — evaluation methodology and results ↗
- MMLU Pro — evaluation methodology and results ↗
- TaxEval — evaluation methodology and results ↗
Names used in source tables
DeepSeek V3 (03/24/2025)