Model comparison
GPT-5 vs GPT-5 Mini
GPT-5 is the stronger model overall, scoring 50.9 to 41.8 on the Noometry Index. GPT-5 Mini costs 5.0× less per token, which makes it the better buy when GPT-5's lead doesn't matter for your workload.
Last verified . 57 shared benchmarks.
Summary
- They share 57 benchmarks with published results for both. GPT-5 scores higher in 9 categories and GPT-5 Mini in 1 category; 10 gaps are clear of the uncertainty.
- The widest gap is in long context, where GPT-5 leads 69.5 to 41.9.
- The biggest single-benchmark swing is SimpleQA Verified: 50.1% for GPT-5 and 21.6% for GPT-5 Mini.
- GPT-5 Mini is cheaper at $0.25 / $2 per million input/output tokens, against $1.25 / $10 for GPT-5.
Side by side
| GPT-5 | GPT-5 Mini | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 50.9 | 41.8 |
| Released | 2025-08-07 | 2025-08-07 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 400K |
| Max output | 128K | 128K |
| Input $ / M tokens | $1.25 | $0.25 |
| Output $ / M tokens | $10 | $2 |
| Results tracked | 69 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 leads
GPT-5: 50.3 (#47), GPT-5 Mini: 40.1 (#146)
| Benchmark | GPT-5 | GPT-5 Mini |
|---|---|---|
| SWE-bench Verified | 73.6% | 64.7% |
| SWE-bench Verified (bash only) | 65% | 59.8% |
| SciCode | 42.9% | 39.2% |
| WeirdML | 60.7% | 52.7% |
| LMArena Coding | 1436 | 1406 |
| ALE-Bench | 1,162 | 799.77 |
| AlgoTune | 1.67 | 1.38 |
| Aider Polyglot | 88% | — |
| LMArena WebDev | 1418 | — |
| SWE-bench Multilingual | — | 39.7% |
| GSO | 6.9% | — |
Agentic & Tool Use GPT-5 leads
GPT-5: 33.1 (#56), GPT-5 Mini: 31.1 (#70)
| Benchmark | GPT-5 | GPT-5 Mini |
|---|---|---|
| Terminal-Bench | 49.6% | 34.8% |
| Berkeley Function Calling Leaderboard | — | 55.5% |
| GDPval | 34.8% | — |
| Remote Labor Index | 1.7% | — |
| DeepResearch Bench | 49.6% | — |
| BALROG | 32.8% | — |
| LMArena Search | 1133 | — |
| METR Time Horizons | 69.6% | — |
| Vending-Bench 2 | — | -31.18 |
Reasoning GPT-5 leads
GPT-5: 38.3 (#64), GPT-5 Mini: 23.9 (#168)
| Benchmark | GPT-5 | GPT-5 Mini |
|---|---|---|
| ARC-AGI-2 | 9.9% | 4.4% |
| Kagi LLM Benchmark | 72.7% | 70.3% |
| ARC-AGI-1 | 65.7% | 54.3% |
| CritPt | 12.6% | 0% |
| Chess Puzzles | 37% | 30% |
| EnigmaEval | 10.5% | 8.2% |
| LMArena Hard Prompts | 1416 | 1380 |
| Mystery Game Puzzles | 23% | 10% |
| DTBench | 90.7% | 80.5% |
| LMCA | 40% | 34.2% |
| Epoch Capabilities Index | 150 | 145.52 |
| ForecastBench | 61.4 | 61 |
| SimpleBench | 56.7% | — |
| EBR-Bench | 12.7% | — |
Math GPT-5 leads
GPT-5: 55.0 (#44), GPT-5 Mini: 46.7 (#69)
| Benchmark | GPT-5 | GPT-5 Mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 55.4% | 46.7% |
| FrontierMath Tier 4 | 22% | 12.2% |
| OTIS Mock AIME 2024-2025 | 91.4% | 86.7% |
| ProofBench | 18% | 9% |
| Omni-MATH | 64.7% | 72.2% |
| LMArena Math | 1407 | 1378 |
| MATH Level 5 | 98.1% | 97.8% |
| FrontierMath (Feb 2025 set) | 32.4% | 27.2% |
| FrontierMath Tier 4 (v1) | 12.5% | 6.3% |
Knowledge GPT-5 leads
GPT-5: 56.6 (#43), GPT-5 Mini: 45.6 (#86)
| Benchmark | GPT-5 | GPT-5 Mini |
|---|---|---|
| GPQA Diamond | 86.2% | 75% |
| Humanity's Last Exam | 25.3% | 19.4% |
| SimpleQA Verified | 50.1% | 21.6% |
| MMLU-Pro | 86.3% | 83.5% |
| Confabulations | 10.3% | 13.3% |
| Vectara Hallucination Rate | 14.7% | 12.9% |
| GPQA (HELM) | 79.2% | 75.6% |
| LMArena Expert | 1419 | 1379 |
Multimodal GPT-5 leads
GPT-5: 46.8 (#13), GPT-5 Mini: 35.6 (#85)
| Benchmark | GPT-5 | GPT-5 Mini |
|---|---|---|
| LMArena Vision | 1232 | 1202 |
| VPCT | 66% | 40.2% |
| GeoBench | 81% | — |
Multilingual GPT-5 leads
GPT-5: 51.4 (#110), GPT-5 Mini: 48.9 (#137)
| Benchmark | GPT-5 | GPT-5 Mini |
|---|---|---|
| LMArena Non-English | 1397 | 1363 |
| LMArena Chinese | 1422 | 1385 |
| LMArena French | 1410 | 1386 |
| LMArena German | 1416 | 1366 |
| LMArena Japanese | 1409 | 1341 |
| LMArena Korean | 1360 | 1308 |
| LMArena Russian | 1406 | 1362 |
| LMArena Spanish | 1399 | 1355 |
Instruction Following GPT-5 Mini leads
GPT-5: 73.8 (#113), GPT-5 Mini: 76.2 (#46)
| Benchmark | GPT-5 | GPT-5 Mini |
|---|---|---|
| IFEval | 87.5% | 92.7% |
| LMArena Instruction Following | 1388 | 1357 |
Long Context GPT-5 leads
GPT-5: 69.5 (#2), GPT-5 Mini: 41.9 (#132)
| Benchmark | GPT-5 | GPT-5 Mini |
|---|---|---|
| Fiction.LiveBench | 97.2% | 69.4% |
| LMArena Longer Query | 1399 | 1355 |
Writing & Preference GPT-5 leads
GPT-5: 63.4 (#65), GPT-5 Mini: 55.2 (#148)
| Benchmark | GPT-5 | GPT-5 Mini |
|---|---|---|
| LMArena Text | 1406 | 1373 |
| LMArena Creative Writing | 1365 | 1325 |
| Short-Story Creative Writing | 86% | 83.1% |
| EQ-Bench Creative Writing | 1627 | 1313 |
| WildBench | 85.7% | 85.5% |
| LMArena Multi-Turn | 1426 | 1363 |
Frequently asked questions
Is GPT-5 better than GPT-5 Mini?
GPT-5 is the stronger model overall, scoring 50.9 to 41.8 on the Noometry Index. GPT-5 Mini costs 5.0× less per token, which makes it the better buy when GPT-5's lead doesn't matter for your workload.
Which is cheaper, GPT-5 or GPT-5 Mini?
GPT-5 Mini is cheaper. It lists at $0.25 per million input tokens and $2 per million output tokens; GPT-5 lists at $1.25 and $10.
Is GPT-5 or GPT-5 Mini better for coding?
GPT-5 scores higher on coding benchmarks: 50.3 versus 40.1 in the Noometry coding category.
Which has the bigger context window?
Both accept 400K tokens.
How many benchmarks do GPT-5 and GPT-5 Mini share?
57 benchmarks have published results for both models. GPT-5 has 69 scored results on Noometry and GPT-5 Mini has 60.