Verdict
DeepSeek-V3 leads 10–0 across 10 shared benchmarks.
DeepSeek-V3 10 · Qwen2.5-72B 0 · higher isn't always better — check the caveats below.
Reported scores — protocols may differ. How we compare
Commonsense Reasoning: DeepSeek-V3 leads 3–0 across 3 shared commonsense reasoning benchmarks (largest gap: Winogrande, 70.4 vs 64.6 — single evaluator).
Reasoning: DeepSeek-V3 leads 3–0 across 3 shared reasoning benchmarks (largest gap: BBH, 83.33 vs 73.07 — single evaluator).
Knowledge: DeepSeek-V3 leads 2–0 across 2 shared knowledge benchmarks (largest gap: TriviaQA, 82.9 vs 71.9 — single evaluator).
Math: DeepSeek-V3 leads 2–0 across 2 shared math benchmarks (largest gap: OTIS Mock AIME 2024-2025, 15.75 vs 7.96 — single evaluator).
Vendor claims — compare against the measured scores above. A claim is what a developer says about its own model, not an independent measurement.
Coding
“DeepSeek-V3 achieves the best performance on most benchmarks, especially on math and code tasks.”
General
“At an economical cost of only 2.664M H800 GPU hours, we complete the pre-training of DeepSeek-V3 on 14.8T tokens, producing the currently strongest open-source base model.”
“DeepSeek-V3 outperforms other open-source models and achieves performance comparable to leading closed-source models.”
Reasoning
“Our pipeline elegantly incorporates the verification and reflection patterns of R1 into DeepSeek-V3 and notably improves its reasoning performance.”
No vendor claims on file for Qwen2.5-72B.
DeepSeek-V3 leads 10–0 across 10 shared benchmarks. DeepSeek-V3 leads 10 benchmarks and Qwen2.5-72B leads 0. Higher isn't always better — see the integrity caveats on each benchmark.
DeepSeek-V3 and Qwen2.5-72B have 10 benchmarks in common in our data — those are the only rows where a direct, apples-to-apples comparison is drawn.
DeepSeek-V3 leads commonsense reasoning 3–0. DeepSeek-V3 leads reasoning 3–0.