Updated daily at 05:00 UTC
Vision-language
GPT-4o mini
- Released
- 18 Jul 2024
- Status
- Legacy
- Context
- 128K
- Max output
- 16.4K
- Input per 1M tokens
- $0.15
- Output per 1M tokens
- $0.60
- Open weights
- No
- License
- proprietary
- Parameters
- —
- Knowledge cutoff
- Oct 2023
- Reasoning
- No
- Modalities
- text, image → text
Category scores
Out of 100, from independent and third-party results only, by method category-2026-09-29.1.
| Category | Score | Rank | Coverage |
|---|---|---|---|
| Knowledge & reasoning | 4.1 | 104 | 3 of 4 |
| Coding & software engineering | 3.6 | Not ranked | 1 of 5 |
| Agentic tool use | 17.4 | Not ranked | 1 of 2 |
| Vision & multimodal understanding | 32.5 | 12 | 2 of 3 |
| Factuality & hallucination | 8.3 | 69 | 1 of 1 |
| Human preference | 26.6 | Not ranked | 2 of 5 |
Benchmark results
Every result we hold for this model, with who ran it and where it was published.
Knowledge & reasoning
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| Chess Puzzles | Puzzles solved | 0.0% | Independent | — | 15 Jul 2026 | Epoch AI |
| DTBench (Decision Theory Benchmark) | Accuracy (capability questions) | 54.4% | Third party | — | — | https://conceptualreasoning.ai/dtbench via Epoch AI |
| LMCA (Language Model Conceptual Argumentation) | Correlation with expert ratings (Pearson r) | 0.1037 | Third party | — | — | https://conceptualreasoning.ai/lmca via Epoch AI |
| MMLU | Accuracy | 76.7% | Third party | — | — | Stanford CRFM Leaderboard via Epoch AI |
| MMLU | Accuracy | 81.8% | Self-reported | — | — | Phi-4 Technical Report via Epoch AI |
| Mystery Game Puzzles | Puzzles solved | 12.0% | Independent | — | 27 Aug 2026 | Epoch AI |
| PIQA | Accuracy | 88.7% | Self-reported | — | — | Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone via Epoch AI |
| SimpleBench | Score (AVG@5) | 10.7% | Third party | — | — | SimpleBench Leaderboard via Epoch AI |
Scientific reasoning
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| GPQA Diamond | Accuracy | 37.7% | Independent | — | 27 Jan 2025 | Epoch AI |
Mathematics
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| FrontierMath Tiers 1-3 | Accuracy | 0.7% | Independent | — | 27 Aug 2026 | Epoch AI |
| GSM8K | Exact match | 91.3% | Self-reported | — | — | Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone via Epoch AI |
| MATH Level 5 | Accuracy | 52.6% | Independent | — | 27 Jan 2025 | Epoch AI |
| OTIS Mock AIME 2024-2025 | Accuracy | 6.9% | Independent | — | 30 Jul 2025 | Epoch AI |
Coding & software engineering
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| Aider Polyglot | Percent correct | 3.6% | Third party | — | — | https://aider.chat/docs/leaderboards/ via Epoch AI |
| WeirdML (v2) | Average accuracy | 11.8% | Third party | — | — | https://htihle.github.io/weirdml.html via Epoch AI |
Agentic tool use
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| BALROG | Average progress | 17.4% | Third party | — | — | https://balrogai.com/ via Epoch AI |
Vision & multimodal understanding
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| GeoBench | Country accuracy | 64.0% | Third party | — | — | GeoBench leaderboard via Epoch AI |
| Visual Physics Comprehension Test (VPCT) | Correct | 34.0% | Third party | — | — | VPCT leaderboard via Epoch AI |
Factuality & hallucination
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| SimpleQA Verified | Accuracy (% correct) | 8.3% | Independent | — | 31 Aug 2026 | Epoch AI |
Writing
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| LLM Creative Story-Writing Benchmark V1 (Lech Mazur) | Mean rubric score (percent of the 10-point scale) | 67.2% | Third party | — | — | lechmazur/writing Github repository via Epoch AI |
Human preference
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| Text Arena (arena.ai) | Arena score (Elo scale) | 1318 | Independent | — | 25 Sep 2026 | Arena |
| Text Arena (arena.ai) | Votes | 68697 | Independent | — | 25 Sep 2026 | Arena |
| Vision Arena (arena.ai) | Arena score (Elo scale) | 1097 | Independent | — | 28 Sep 2026 | Arena |
| Vision Arena (arena.ai) | Votes | 17341 | Independent | — | 28 Sep 2026 | Arena |
Other GPT-4o versions
- GPT-4o mini Search Preview11 Mar 2025
- GPT-4o Search Preview11 Mar 2025
- ChatGPT-4o15 Aug 2024
- GPT-4o13 May 2024