GPT-4o is the stronger all-rounder.
Scores updated · 24 tests both models report · How we compare
Where each one wins
Tests won in each of the two areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
Reasoning and long documents rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where GPT-4o pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+17.3points ahead
- Questions about very long textsLongBench v2+16.5points ahead
Where DeepSeek-V2 pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 24 tests both models report. The winning score is in its model's colour; marks a score checked independently.
Other results22 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- CodeforcesGPT-4o by 741.5 rating points17.5759+741.5 rating
- HumanEvalGPT-4o by 46.943.390.2+46.9
- MATHGPT-4o by 41.943.485.3+41.9
- MGSMGPT-4o by 26.963.690.5+26.9
- LiveCodeBenchGPT-4o by 26.711.638.3+26.7
- IFEvalGPT-4o by 23.357.781+23.3
- MMLU-ProGPT-4o by 23.351.474.7+23.3
- MATH-500 (EM)GPT-4o by 18.356.374.6+18.3
- MMMLUGPT-4o by 17.46481.4+17.4
- FRAMES (Acc.)GPT-4o by 13.666.980.5+13.6
- Aider-Edit (Acc.)GPT-4o by 12.660.372.9+12.6
- GSM8KGPT-4o by 11.779.290.9+11.7
- HumanEval-Mul (Pass@1)GPT-4o by 11.269.380.5+11.2
- Chinese SimpleQA (C-SimpleQA)GPT-4o by 10.248.558.7+10.2
- MMLU-ReduxGPT-4o by 10.177.988+10.1
- DROP (F1)GPT-4o by 8.880.489.2+8.8
- CLUEWSCGPT-4o by 5.98287.9+5.9
- C-EvalDeepSeek-V2 by 5.481.476+5.4
- AIME 2024GPT-4o by 4.74.69.3+4.7
- DROPGPT-4o by 3.380.183.4+3.3
- MMLUDeepSeek-V2 by 1.875.673.8+1.8
- DROP (3-shot F1)tie8383.7tie
Questions people ask
Which is better, DeepSeek-V2 or GPT-4o?
GPT-4o wins all two areas where both have results: reasoning and long documents. DeepSeek-V2 wins none.
How do you compare the two?
We use the 24 benchmark tests both models have published scores on. The verdict counts the 2 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 22 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.