Model comparison
GPT-5.4 mini vs GPT-5 Mini
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 41.8 on the Noometry Index. GPT-5 Mini costs 2.5× less per token, which makes it the better buy when GPT-5.4 mini's lead doesn't matter for your workload.
Last verified . 41 shared benchmarks.
Summary
- They share 41 benchmarks with published results for both. GPT-5.4 mini scores higher in 7 categories and GPT-5 Mini in 3 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GPT-5.4 mini leads 64.0 to 55.2.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 37.9% for GPT-5.4 mini and 70.3% for GPT-5 Mini.
- GPT-5 Mini is cheaper at $0.25 / $2 per million input/output tokens, against $0.75 / $4.50 for GPT-5.4 mini.
Side by side
| GPT-5.4 mini | GPT-5 Mini | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 45.0 | 41.8 |
| Released | 2026-03-17 | 2025-08-07 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 400K |
| Max output | 128K | 128K |
| Input $ / M tokens | $0.75 | $0.25 |
| Output $ / M tokens | $4.50 | $2 |
| Results tracked | 46 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 mini leads
GPT-5.4 mini: 45.2 (#72), GPT-5 Mini: 40.1 (#146)
| Benchmark | GPT-5.4 mini | GPT-5 Mini |
|---|---|---|
| SciCode | 49.9% | 39.2% |
| WeirdML | 60.3% | 52.7% |
| LMArena Coding | 1438 | 1406 |
| ALE-Bench | 1,189 | 799.77 |
| SWE-bench Verified | — | 64.7% |
| FrontierCode | 27% | — |
| SWE-bench Verified (bash only) | — | 59.8% |
| LMArena WebDev | 1397 | — |
| SWE-bench Multilingual | — | 39.7% |
| AlgoTune | — | 1.38 |
Agentic & Tool Use GPT-5 Mini leads
GPT-5.4 mini: 29.9 (#81), GPT-5 Mini: 31.1 (#70)
| Benchmark | GPT-5.4 mini | GPT-5 Mini |
|---|---|---|
| Terminal-Bench | — | 34.8% |
| Berkeley Function Calling Leaderboard | — | 55.5% |
| DeepResearch Bench | 36.3% | — |
| Vending-Bench 2 | — | -31.18 |
Reasoning GPT-5.4 mini leads
GPT-5.4 mini: 30.4 (#85), GPT-5 Mini: 23.9 (#168)
| Benchmark | GPT-5.4 mini | GPT-5 Mini |
|---|---|---|
| ARC-AGI-2 | 18.9% | 4.4% |
| Kagi LLM Benchmark | 37.9% | 70.3% |
| ARC-AGI-1 | 63.7% | 54.3% |
| CritPt | 10% | 0% |
| Chess Puzzles | 24% | 30% |
| LMArena Hard Prompts | 1424 | 1380 |
| Mystery Game Puzzles | 11% | 10% |
| DTBench | 80% | 80.5% |
| LMCA | 40.8% | 34.2% |
| Epoch Capabilities Index | 148.84 | 145.52 |
| ForecastBench | 57 | 61 |
| NYT Connections (extended) | 61.8% | — |
| EnigmaEval | — | 8.2% |
| Thematic Generalization | 61.7% | — |
Math GPT-5 Mini leads
GPT-5.4 mini: 45.5 (#75), GPT-5 Mini: 46.7 (#69)
| Benchmark | GPT-5.4 mini | GPT-5 Mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 51.2% | 46.7% |
| FrontierMath Tier 4 | 9.8% | 12.2% |
| OTIS Mock AIME 2024-2025 | 88.9% | 86.7% |
| ProofBench | 21% | 9% |
| LMArena Math | 1419 | 1378 |
| FrontierMath (Feb 2025 set) | 28.3% | 27.2% |
| FrontierMath Tier 4 (v1) | 2.1% | 6.3% |
| Omni-MATH | — | 72.2% |
| MATH Level 5 | — | 97.8% |
Knowledge GPT-5.4 mini leads
GPT-5.4 mini: 51.5 (#67), GPT-5 Mini: 45.6 (#86)
| Benchmark | GPT-5.4 mini | GPT-5 Mini |
|---|---|---|
| GPQA Diamond | 86.9% | 75% |
| SimpleQA Verified | 29.4% | 21.6% |
| Vectara Hallucination Rate | 5.5% | 12.9% |
| LMArena Expert | 1435 | 1379 |
| Humanity's Last Exam | — | 19.4% |
| MMLU-Pro | — | 83.5% |
| Confabulations | — | 13.3% |
| GPQA (HELM) | — | 75.6% |
Multimodal GPT-5.4 mini leads
GPT-5.4 mini: 39.7 (#56), GPT-5 Mini: 35.6 (#85)
| Benchmark | GPT-5.4 mini | GPT-5 Mini |
|---|---|---|
| LMArena Vision | 1245 | 1202 |
| VPCT | — | 40.2% |
Multilingual GPT-5.4 mini leads
GPT-5.4 mini: 51.9 (#96), GPT-5 Mini: 48.9 (#137)
| Benchmark | GPT-5.4 mini | GPT-5 Mini |
|---|---|---|
| LMArena Non-English | 1405 | 1363 |
| LMArena Chinese | 1446 | 1385 |
| LMArena French | 1440 | 1386 |
| LMArena German | 1409 | 1366 |
| LMArena Japanese | 1374 | 1341 |
| LMArena Korean | 1368 | 1308 |
| LMArena Russian | 1417 | 1362 |
| LMArena Spanish | 1405 | 1355 |
Instruction Following GPT-5 Mini leads
GPT-5.4 mini: 74.1 (#102), GPT-5 Mini: 76.2 (#46)
| Benchmark | GPT-5.4 mini | GPT-5 Mini |
|---|---|---|
| LMArena Instruction Following | 1405 | 1357 |
| IFEval | — | 92.7% |
Long Context GPT-5.4 mini leads
GPT-5.4 mini: 43.0 (#112), GPT-5 Mini: 41.9 (#132)
| Benchmark | GPT-5.4 mini | GPT-5 Mini |
|---|---|---|
| LMArena Longer Query | 1407 | 1355 |
| Fiction.LiveBench | — | 69.4% |
Writing & Preference GPT-5.4 mini leads
GPT-5.4 mini: 64.0 (#58), GPT-5 Mini: 55.2 (#148)
| Benchmark | GPT-5.4 mini | GPT-5 Mini |
|---|---|---|
| LMArena Text | 1412 | 1373 |
| LMArena Creative Writing | 1370 | 1325 |
| EQ-Bench Creative Writing | 1665 | 1313 |
| LMArena Multi-Turn | 1429 | 1363 |
| Short-Story Creative Writing | — | 83.1% |
| WildBench | — | 85.5% |
Frequently asked questions
Is GPT-5.4 mini better than GPT-5 Mini?
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 41.8 on the Noometry Index. GPT-5 Mini costs 2.5× less per token, which makes it the better buy when GPT-5.4 mini's lead doesn't matter for your workload.
Which is cheaper, GPT-5.4 mini 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-5.4 mini lists at $0.75 and $4.50.
Is GPT-5.4 mini or GPT-5 Mini better for coding?
GPT-5.4 mini scores higher on coding benchmarks: 45.2 versus 40.1 in the Noometry coding category.
Which has the bigger context window?
Both accept 400K tokens.
How many benchmarks do GPT-5.4 mini and GPT-5 Mini share?
41 benchmarks have published results for both models. GPT-5.4 mini has 46 scored results on Noometry and GPT-5 Mini has 60.