Model comparison
GPT-4 Turbo vs GPT-5 Mini
GPT-5 Mini is the stronger model overall, scoring 41.7 to 30.3 on the Noometry Index.
Last verified . 29 shared benchmarks.
Summary
- They share 29 benchmarks with published results for both. GPT-4 Turbo scores higher in 0 categories and GPT-5 Mini in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5 Mini leads 46.3 to 8.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 6.7% for GPT-4 Turbo and 86.7% for GPT-5 Mini.
- GPT-5 Mini is cheaper at $0.25 / $2 per million input/output tokens, against $10 / $30 for GPT-4 Turbo.
- GPT-5 Mini accepts more context: 400K tokens versus 128K.
Side by side
| GPT-4 Turbo | GPT-5 Mini | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 30.3 | 41.7 |
| Released | 2023-11-06 | 2025-08-07 |
| Weights | Proprietary | Proprietary |
| Context window | 128K | 400K |
| Max output | 4K | 128K |
| Input $ / M tokens | $10 | $0.25 |
| Output $ / M tokens | $30 | $2 |
| Results tracked | 36 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 Mini leads
GPT-4 Turbo: 33.4 (#247), GPT-5 Mini: 40.0 (#134)
| Benchmark | GPT-4 Turbo | GPT-5 Mini |
|---|---|---|
| WeirdML | 18% | 52.7% |
| LMArena Coding | 1268 | 1406 |
| SWE-bench Verified | — | 64.7% |
| SWE-bench Verified (bash only) | — | 59.8% |
| SWE-bench Multilingual | — | 39.7% |
| SciCode | — | 39.2% |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 58.2% | — |
| ALE-Bench | — | 799.77 |
| AlgoTune | — | 1.38 |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73.3% | — |
Agentic & Tool Use Not comparable
GPT-4 Turbo: —, GPT-5 Mini: 31.0 (#64)
| Benchmark | GPT-4 Turbo | GPT-5 Mini |
|---|---|---|
| Terminal-Bench | — | 34.8% |
| Berkeley Function Calling Leaderboard | — | 55.5% |
| METR Time Horizons | 36.7% | — |
| Vending-Bench 2 | — | -31.18 |
Reasoning GPT-5 Mini leads
GPT-4 Turbo: 14.9 (#311), GPT-5 Mini: 24.1 (#153)
| Benchmark | GPT-4 Turbo | GPT-5 Mini |
|---|---|---|
| Chess Puzzles | 6% | 30% |
| LMArena Hard Prompts | 1251 | 1380 |
| DTBench | 61.6% | 80.5% |
| LMCA | 9.8% | 34.2% |
| Epoch Capabilities Index | 127.25 | 145.51 |
| ForecastBench | 59.4 | 61 |
| ARC-AGI-2 | — | 4.4% |
| SimpleBench | 25.1% | — |
| Kagi LLM Benchmark | — | 70.3% |
| ARC-AGI-1 | — | 54.3% |
| CritPt | — | 0% |
| EnigmaEval | — | 8.2% |
| Mystery Game Puzzles | — | 10% |
Math GPT-5 Mini leads
GPT-4 Turbo: 8.6 (#316), GPT-5 Mini: 46.3 (#59)
| Benchmark | GPT-4 Turbo | GPT-5 Mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 0.7% | 46.7% |
| OTIS Mock AIME 2024-2025 | 6.7% | 86.7% |
| LMArena Math | 1272 | 1377 |
| MATH Level 5 | 46.7% | 97.8% |
| FrontierMath Tier 4 | — | 12.2% |
| ProofBench | — | 9% |
| Omni-MATH | — | 72.2% |
| FrontierMath (Feb 2025 set) | — | 27.2% |
| FrontierMath Tier 4 (v1) | — | 6.3% |
Knowledge GPT-5 Mini leads
GPT-4 Turbo: 24.3 (#263), GPT-5 Mini: 45.5 (#80)
| Benchmark | GPT-4 Turbo | GPT-5 Mini |
|---|---|---|
| GPQA Diamond | 46.6% | 75% |
| Confabulations | 28.4% | 13.3% |
| LMArena Expert | 1223 | 1377 |
| Humanity's Last Exam | — | 19.4% |
| SimpleQA Verified | — | 21.6% |
| MMLU-Pro | — | 83.5% |
| Vectara Hallucination Rate | — | 12.9% |
| GPQA (HELM) | — | 75.6% |
| MMLU | 81.3% | — |
Multimodal GPT-5 Mini leads
GPT-4 Turbo: 30.0 (#104), GPT-5 Mini: 35.0 (#80)
| Benchmark | GPT-4 Turbo | GPT-5 Mini |
|---|---|---|
| LMArena Vision | 1090 | 1202 |
| VPCT | — | 40.2% |
Multilingual GPT-5 Mini leads
GPT-4 Turbo: 40.7 (#211), GPT-5 Mini: 49.1 (#131)
| Benchmark | GPT-4 Turbo | GPT-5 Mini |
|---|---|---|
| LMArena Non-English | 1245 | 1363 |
| LMArena Chinese | 1242 | 1387 |
| LMArena French | 1275 | 1385 |
| LMArena German | 1259 | 1366 |
| LMArena Japanese | 1194 | 1342 |
| LMArena Korean | 1187 | 1306 |
| LMArena Russian | 1259 | 1362 |
| LMArena Spanish | 1260 | 1355 |
Instruction Following GPT-5 Mini leads
GPT-4 Turbo: 66.0 (#211), GPT-5 Mini: 76.4 (#43)
| Benchmark | GPT-4 Turbo | GPT-5 Mini |
|---|---|---|
| LMArena Instruction Following | 1250 | 1357 |
| IFEval | — | 92.7% |
Long Context GPT-5 Mini leads
GPT-4 Turbo: 38.0 (#200), GPT-5 Mini: 41.9 (#127)
| Benchmark | GPT-4 Turbo | GPT-5 Mini |
|---|---|---|
| LMArena Longer Query | 1254 | 1355 |
| Fiction.LiveBench | — | 69.4% |
Writing & Preference GPT-5 Mini leads
GPT-4 Turbo: 47.9 (#201), GPT-5 Mini: 55.4 (#143)
| Benchmark | GPT-4 Turbo | GPT-5 Mini |
|---|---|---|
| LMArena Text | 1272 | 1373 |
| LMArena Creative Writing | 1269 | 1325 |
| LMArena Multi-Turn | 1267 | 1363 |
| Short-Story Creative Writing | — | 83.1% |
| EQ-Bench Creative Writing | — | 1313 |
| WildBench | — | 85.5% |
Frequently asked questions
Is GPT-4 Turbo better than GPT-5 Mini?
GPT-5 Mini is the stronger model overall, scoring 41.7 to 30.3 on the Noometry Index.
Which is cheaper, GPT-4 Turbo 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-4 Turbo lists at $10 and $30.
Is GPT-4 Turbo or GPT-5 Mini better for coding?
GPT-5 Mini scores higher on coding benchmarks: 40.0 versus 33.4 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Mini does, with 400K tokens against 128K.
How many benchmarks do GPT-4 Turbo and GPT-5 Mini share?
29 benchmarks have published results for both models. GPT-4 Turbo has 36 scored results on Noometry and GPT-5 Mini has 60.