Model comparison
GPT-4o mini vs GPT-5
GPT-5 is the stronger model overall, scoring 50.9 to 25.4 on the Noometry Index. GPT-4o mini costs 13× less per token, which makes it the better buy when GPT-5's lead doesn't matter for your workload.
Last verified . 44 shared benchmarks.
Summary
- They share 44 benchmarks with published results for both. GPT-4o mini scores higher in 0 categories and GPT-5 in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5 leads 54.6 to 9.9.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 6.9% for GPT-4o mini and 91.4% for GPT-5.
- GPT-4o mini is cheaper at $0.15 / $0.60 per million input/output tokens, against $1.25 / $10 for GPT-5.
- GPT-5 accepts more context: 400K tokens versus 128K.
Side by side
| GPT-4o mini | GPT-5 | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 25.4 | 50.9 |
| Released | 2024-07-18 | 2025-08-07 |
| Weights | Proprietary | Proprietary |
| Context window | 128K | 400K |
| Max output | 16K | 128K |
| Input $ / M tokens | $0.15 | $1.25 |
| Output $ / M tokens | $0.60 | $10 |
| Results tracked | 60 | 69 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 leads
GPT-4o mini: 21.6 (#330), GPT-5: 50.7 (#39)
| Benchmark | GPT-4o mini | GPT-5 |
|---|---|---|
| Aider Polyglot | 3.6% | 88% |
| WeirdML | 11.8% | 60.7% |
| LMArena Coding | 1290 | 1436 |
| SWE-bench Verified | — | 73.6% |
| SWE-bench Verified (bash only) | — | 65% |
| LMArena WebDev | — | 1417 |
| SciCode | — | 42.9% |
| GSO | — | 6.9% |
| BigCodeBench Instruct | 46.1% | — |
| LiveBench Coding | 43.1% | — |
| BigCodeBench Complete | 57.4% | — |
| ALE-Bench | — | 1,162 |
| AlgoTune | — | 1.67 |
| HumanEval+ | 83.5% | — |
| MBPP+ | 72.2% | — |
Agentic & Tool Use GPT-5 leads
GPT-4o mini: 27.2 (#95), GPT-5: 33.0 (#53)
| Benchmark | GPT-4o mini | GPT-5 |
|---|---|---|
| BALROG | 17.4% | 32.8% |
| Terminal-Bench | — | 49.6% |
| GDPval | — | 34.8% |
| Remote Labor Index | — | 1.7% |
| DeepResearch Bench | — | 49.6% |
| LMArena Search | — | 1133 |
| METR Time Horizons | — | 69.6% |
Reasoning GPT-5 leads
GPT-4o mini: 8.7 (#343), GPT-5: 38.3 (#55)
| Benchmark | GPT-4o mini | GPT-5 |
|---|---|---|
| ARC-AGI-2 | 0% | 9.9% |
| SimpleBench | 10.7% | 56.7% |
| Kagi LLM Benchmark | 28.8% | 72.7% |
| Chess Puzzles | 0% | 37% |
| LMArena Hard Prompts | 1267 | 1415 |
| Mystery Game Puzzles | 12% | 23% |
| DTBench | 54.4% | 90.7% |
| LMCA | 10.4% | 40% |
| Epoch Capabilities Index | 126.56 | 150 |
| ARC-AGI-1 | — | 65.7% |
| CritPt | — | 12.6% |
| EnigmaEval | — | 10.5% |
| EBR-Bench | — | 12.7% |
| LiveBench Reasoning | 32.8% | — |
| LiveBench Data Analysis | 50% | — |
| ForecastBench | — | 61.4 |
| LiveBench | 41.3% | — |
| PIQA | 88.7% | — |
Math GPT-5 leads
GPT-4o mini: 9.9 (#307), GPT-5: 54.6 (#36)
| Benchmark | GPT-4o mini | GPT-5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 0.7% | 55.4% |
| OTIS Mock AIME 2024-2025 | 6.9% | 91.4% |
| Omni-MATH | 28% | 64.7% |
| LMArena Math | 1267 | 1406 |
| MATH Level 5 | 52.6% | 98.1% |
| FrontierMath Tier 4 | — | 22% |
| ProofBench | — | 18% |
| LiveBench Math | 36.3% | — |
| FrontierMath (Feb 2025 set) | — | 32.4% |
| FrontierMath Tier 4 (v1) | — | 12.5% |
| GSM8K | 91.3% | — |
Knowledge GPT-5 leads
GPT-4o mini: 17.7 (#279), GPT-5: 56.5 (#37)
| Benchmark | GPT-4o mini | GPT-5 |
|---|---|---|
| GPQA Diamond | 37.7% | 86.2% |
| SimpleQA Verified | 8.3% | 50.1% |
| MMLU-Pro | 60.3% | 86.3% |
| Confabulations | 37.2% | 10.3% |
| GPQA (HELM) | 36.8% | 79.2% |
| LMArena Expert | 1235 | 1420 |
| Humanity's Last Exam | — | 25.3% |
| Vectara Hallucination Rate | — | 14.7% |
| BoolQ | 88.7% | — |
| MMLU | 81.8% | — |
Multimodal GPT-5 leads
GPT-4o mini: 25.3 (#117), GPT-5: 46.1 (#12)
| Benchmark | GPT-4o mini | GPT-5 |
|---|---|---|
| LMArena Vision | 1066 | 1232 |
| GeoBench | 64% | 81% |
| VPCT | 34% | 66% |
| Video-MME | 64.8% | — |
Multilingual GPT-5 leads
GPT-4o mini: 42.2 (#194), GPT-5: 51.5 (#105)
| Benchmark | GPT-4o mini | GPT-5 |
|---|---|---|
| LMArena Non-English | 1266 | 1397 |
| LMArena Chinese | 1264 | 1423 |
| LMArena French | 1296 | 1409 |
| LMArena German | 1272 | 1415 |
| LMArena Japanese | 1216 | 1411 |
| LMArena Korean | 1195 | 1361 |
| LMArena Russian | 1275 | 1406 |
| LMArena Spanish | 1276 | 1398 |
Instruction Following GPT-5 leads
GPT-4o mini: 62.1 (#234), GPT-5: 74.0 (#107)
| Benchmark | GPT-4o mini | GPT-5 |
|---|---|---|
| IFEval | 78.2% | 87.5% |
| LMArena Instruction Following | 1259 | 1388 |
| LiveBench Instruction Following | 56.8% | — |
Long Context GPT-5 leads
GPT-4o mini: 39.1 (#181), GPT-5: 69.5 (#2)
| Benchmark | GPT-4o mini | GPT-5 |
|---|---|---|
| LMArena Longer Query | 1289 | 1399 |
| Fiction.LiveBench | — | 97.2% |
Writing & Preference GPT-5 leads
GPT-4o mini: 39.7 (#243), GPT-5: 63.5 (#63)
| Benchmark | GPT-4o mini | GPT-5 |
|---|---|---|
| LMArena Text | 1286 | 1406 |
| LMArena Creative Writing | 1268 | 1365 |
| Short-Story Creative Writing | 67.2% | 86% |
| EQ-Bench Creative Writing | 873 | 1627 |
| WildBench | 79.1% | 85.7% |
| LMArena Multi-Turn | 1285 | 1426 |
| LiveBench Language | 28.6% | — |
Frequently asked questions
Is GPT-4o mini better than GPT-5?
GPT-5 is the stronger model overall, scoring 50.9 to 25.4 on the Noometry Index. GPT-4o mini costs 13× less per token, which makes it the better buy when GPT-5's lead doesn't matter for your workload.
Which is cheaper, GPT-4o mini or GPT-5?
GPT-4o mini is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; GPT-5 lists at $1.25 and $10.
Is GPT-4o mini or GPT-5 better for coding?
GPT-5 scores higher on coding benchmarks: 50.7 versus 21.6 in the Noometry coding category.
Which has the bigger context window?
GPT-5 does, with 400K tokens against 128K.
How many benchmarks do GPT-4o mini and GPT-5 share?
44 benchmarks have published results for both models. GPT-4o mini has 60 scored results on Noometry and GPT-5 has 69.