Compare / head-to-head
GPT-4o minivsGPT-5
GPT-5 leads 12 of 12 shared benchmarks. GPT-4o mini is 15x cheaper per token. GPT-5 has the larger context window (400K tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
0 – 12
GPT-5 leads
Cheaper per token
GPT-4o mini
15x cheaper, input + output
Larger context
GPT-5
400K tokens
Providers
OpenAI
same provider
GPT-4o mini
OpenAI · released 2024-07-18
textvision
Quality
Benchmark matrix
12 shared · 2 only GPT-4o mini · 8 only GPT-5| Benchmark | GPT-4o mini | GPT-5 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 12 | ||||
| GPQA Diamond | 40.2% ↗ | 86.2% ↗epoch run | -46 pt | GPT-5 |
| LMArena Elo | 1317.6 ↗ | 1434.5 ↗ | -116.9 | GPT-5 |
| BALROG (BALROG) | 17.4% ↗ | 32.8% ↗ | -15.4 pt | GPT-5 |
| Chess Puzzles (Epoch AI run) | 0% ↗epoch run | 37% ↗epoch run | -37 pt | GPT-5 |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 0.7% ↗epoch run | 55.4% ↗epoch run | -54.7 pt | GPT-5 |
| LMArena Vision (LMArena) | 1097.1 ↗ | 1210.2 ↗ | -113.1 | GPT-5 |
| MATH Level 5 (Epoch AI run) | 52.6% ↗epoch run | 98.1% ↗epoch run | -45.5 pt | GPT-5 |
| Mystery Game Puzzles (Epoch AI run) | 12% ↗epoch run | 23% ↗epoch run | -11 pt | GPT-5 |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 6.9% ↗epoch run | 91.4% ↗epoch run | -84.5 pt | GPT-5 |
| SimpleBench (SimpleBench) | 10.7% ↗ | 56.7% ↗ | -46 pt | GPT-5 |
| SimpleQA Verified | 8.3% ↗epoch run | 50.1% ↗epoch run | -41.8 pt | GPT-5 |
| WeirdML (Håvard Tveit Ihle) | 11.8% ↗ | 60.7% ↗ | -48.9 pt | GPT-5 |
| Only GPT-4o mini reports · 2 | ||||
| MMLU | 82% ↗ | not reported | — | — |
| AIME 2024 | 8.6% ↗ | not reported | — | — |
| Only GPT-5 reports · 8 | ||||
| SWE-bench Verified | not reported | 74.9% ↗ | — | — |
| AIME 2025 | not reported | 94.6% ↗ | — | — |
| EBR-bench (Epoch AI run) | not reported | 12.7% ↗epoch run | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | not reported | 22% ↗epoch run | — | — |
| GSO Opt@1 (GSO) | not reported | 5.9% ↗ | — | — |
| LMArena WebDev (LMArena) | not reported | 1417.1 ↗ | — | — |
| SAGE (Vals AI) | not reported | 43.7% ↗ | — | — |
| SWE-bench Verified (Epoch AI run) | not reported | 73.6% ↗epoch run | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is GPT-4o mini minus GPT-5 in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | GPT-4o mini | GPT-5 | Edge |
|---|---|---|---|
| Context window | 128K tokens | 400K tokens | GPT-5 |
| Max output | 16K tokens | 128K tokens | GPT-5 |
| Input price / 1M | $0.15 ↗ | $1.25 ↗ | GPT-4o mini |
| Output price / 1M | $0.6 ↗ | $10 ↗ | GPT-4o mini |
| Cached input / 1M | $0.075 ↗ | $0.125 ↗ | GPT-4o mini |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $0.75 | $11.25 | GPT-4o mini |
| Modalities | text · vision | text · vision | Tie |
| Released | 2024-07-18 | 2025-08-07 | — |
| Cited benchmark scores | 19 | 27 | — |
Reliability
OpenAI status
Live from /statusMore matchups
GPT-4o mini vs …
Models sharing the most benchmarksMore matchups
GPT-5 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run GPT-4o mini and GPT-5 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.