Model comparison
GPT-5 Mini vs o4-mini
GPT-5 Mini and o4-mini score almost the same on the Noometry Index (41.8 vs 41.6), so choose on price, context window or the category you care about most.
Last verified . 53 shared benchmarks.
Summary
- They share 53 benchmarks with published results for both. GPT-5 Mini scores higher in 5 categories and o4-mini in 5 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5 Mini leads 46.7 to 40.8.
- The biggest single-benchmark swing is VPCT: 40.2% for GPT-5 Mini and 57.5% for o4-mini.
- GPT-5 Mini is cheaper at $0.25 / $2 per million input/output tokens, against $1.10 / $4.40 for o4-mini.
- GPT-5 Mini accepts more context: 400K tokens versus 200K.
Side by side
| GPT-5 Mini | o4-mini | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 41.8 | 41.6 |
| Released | 2025-08-07 | 2025-04-16 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 200K |
| Max output | 128K | 100K |
| Input $ / M tokens | $0.25 | $1.10 |
| Output $ / M tokens | $2 | $4.40 |
| Results tracked | 60 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GPT-5 Mini: 40.1 (#146), o4-mini: 40.9 (#127)
| Benchmark | GPT-5 Mini | o4-mini |
|---|---|---|
| SWE-bench Verified (bash only) | 59.8% | 45% |
| WeirdML | 52.7% | 52.6% |
| LMArena Coding | 1406 | 1368 |
| ALE-Bench | 799.77 | 826.17 |
| AlgoTune | 1.38 | 1.72 |
| SWE-bench Verified | 64.7% | — |
| Aider Polyglot | — | 72% |
| SWE-bench Multilingual | 39.7% | — |
| SciCode | 39.2% | — |
| GSO | — | 3.6% |
| CadEval | — | 62% |
Agentic & Tool Use o4-mini leads
GPT-5 Mini: 31.1 (#70), o4-mini: 32.6 (#61)
| Benchmark | GPT-5 Mini | o4-mini |
|---|---|---|
| Berkeley Function Calling Leaderboard | 55.5% | 53.2% |
| Terminal-Bench | 34.8% | — |
| GDPval | — | 25.3% |
| METR Time Horizons | — | 63.9% |
| Vending-Bench 2 | -31.18 | — |
Reasoning Too close to call
GPT-5 Mini: 23.9 (#168), o4-mini: 24.6 (#162)
| Benchmark | GPT-5 Mini | o4-mini |
|---|---|---|
| ARC-AGI-2 | 4.4% | 6.1% |
| Kagi LLM Benchmark | 70.3% | 67.6% |
| ARC-AGI-1 | 54.3% | 58.7% |
| CritPt | 0% | 0.6% |
| Chess Puzzles | 30% | 26% |
| EnigmaEval | 8.2% | 9.2% |
| LMArena Hard Prompts | 1380 | 1351 |
| Mystery Game Puzzles | 10% | 5% |
| DTBench | 80.5% | 77.6% |
| LMCA | 34.2% | 26.5% |
| Epoch Capabilities Index | 145.52 | 145.64 |
| ForecastBench | 61 | 61.8 |
| SimpleBench | — | 38.7% |
Math GPT-5 Mini leads
GPT-5 Mini: 46.7 (#69), o4-mini: 40.8 (#89)
| Benchmark | GPT-5 Mini | o4-mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 46.7% | 36.1% |
| FrontierMath Tier 4 | 12.2% | 4.9% |
| OTIS Mock AIME 2024-2025 | 86.7% | 81.7% |
| Omni-MATH | 72.2% | 72% |
| LMArena Math | 1378 | 1389 |
| MATH Level 5 | 97.8% | 97.8% |
| FrontierMath (Feb 2025 set) | 27.2% | 24.8% |
| FrontierMath Tier 4 (v1) | 6.3% | 6.3% |
| ProofBench | 9% | — |
Knowledge GPT-5 Mini leads
GPT-5 Mini: 45.6 (#86), o4-mini: 43.6 (#91)
| Benchmark | GPT-5 Mini | o4-mini |
|---|---|---|
| GPQA Diamond | 75% | 79.6% |
| Humanity's Last Exam | 19.4% | 18.1% |
| SimpleQA Verified | 21.6% | 19.6% |
| MMLU-Pro | 83.5% | 82% |
| Confabulations | 13.3% | 15.8% |
| Vectara Hallucination Rate | 12.9% | 18.6% |
| GPQA (HELM) | 75.6% | 73.5% |
| LMArena Expert | 1379 | 1343 |
Multimodal o4-mini leads
GPT-5 Mini: 35.6 (#85), o4-mini: 40.2 (#49)
| Benchmark | GPT-5 Mini | o4-mini |
|---|---|---|
| LMArena Vision | 1202 | 1194 |
| VPCT | 40.2% | 57.5% |
| GeoBench | — | 64% |
Multilingual GPT-5 Mini leads
GPT-5 Mini: 48.9 (#137), o4-mini: 47.0 (#154)
| Benchmark | GPT-5 Mini | o4-mini |
|---|---|---|
| LMArena Non-English | 1363 | 1337 |
| LMArena Chinese | 1385 | 1354 |
| LMArena French | 1386 | 1364 |
| LMArena German | 1366 | 1336 |
| LMArena Japanese | 1341 | 1308 |
| LMArena Korean | 1308 | 1312 |
| LMArena Russian | 1362 | 1334 |
| LMArena Spanish | 1355 | 1347 |
Instruction Following GPT-5 Mini leads
GPT-5 Mini: 76.2 (#46), o4-mini: 75.2 (#68)
| Benchmark | GPT-5 Mini | o4-mini |
|---|---|---|
| IFEval | 92.7% | 92.8% |
| LMArena Instruction Following | 1357 | 1321 |
Long Context o4-mini leads
GPT-5 Mini: 41.9 (#132), o4-mini: 45.5 (#33)
| Benchmark | GPT-5 Mini | o4-mini |
|---|---|---|
| Fiction.LiveBench | 69.4% | 77.8% |
| LMArena Longer Query | 1355 | 1315 |
Writing & Preference GPT-5 Mini leads
GPT-5 Mini: 55.2 (#148), o4-mini: 54.0 (#152)
| Benchmark | GPT-5 Mini | o4-mini |
|---|---|---|
| LMArena Text | 1373 | 1353 |
| LMArena Creative Writing | 1325 | 1294 |
| Short-Story Creative Writing | 83.1% | 75% |
| WildBench | 85.5% | 85.4% |
| LMArena Multi-Turn | 1363 | 1350 |
| EQ-Bench Creative Writing | 1313 | — |
Frequently asked questions
Is GPT-5 Mini better than o4-mini?
GPT-5 Mini and o4-mini score almost the same on the Noometry Index (41.8 vs 41.6), so choose on price, context window or the category you care about most.
Which is cheaper, GPT-5 Mini or o4-mini?
GPT-5 Mini is cheaper. It lists at $0.25 per million input tokens and $2 per million output tokens; o4-mini lists at $1.10 and $4.40.
Is GPT-5 Mini or o4-mini better for coding?
They score almost the same on coding (40.1 vs 40.9); test both on your own repository before choosing.
Which has the bigger context window?
GPT-5 Mini does, with 400K tokens against 200K.
How many benchmarks do GPT-5 Mini and o4-mini share?
53 benchmarks have published results for both models. GPT-5 Mini has 60 scored results on Noometry and o4-mini has 60.