Model comparison
GPT-5.4 mini vs o4-mini
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 41.6 on the Noometry Index.
Last verified . 38 shared benchmarks.
Summary
- They share 38 benchmarks with published results for both. GPT-5.4 mini scores higher in 6 categories and o4-mini in 4 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GPT-5.4 mini leads 64.0 to 54.0.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 37.9% for GPT-5.4 mini and 67.6% for o4-mini.
- GPT-5.4 mini is cheaper at $0.75 / $4.50 per million input/output tokens, against $1.10 / $4.40 for o4-mini.
- GPT-5.4 mini accepts more context: 400K tokens versus 200K.
Side by side
| GPT-5.4 mini | o4-mini | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 45.0 | 41.6 |
| Released | 2026-03-17 | 2025-04-16 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 200K |
| Max output | 128K | 100K |
| Input $ / M tokens | $0.75 | $1.10 |
| Output $ / M tokens | $4.50 | $4.40 |
| 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), o4-mini: 40.9 (#127)
| Benchmark | GPT-5.4 mini | o4-mini |
|---|---|---|
| WeirdML | 60.3% | 52.6% |
| LMArena Coding | 1438 | 1368 |
| ALE-Bench | 1,189 | 826.17 |
| FrontierCode | 27% | — |
| SWE-bench Verified (bash only) | — | 45% |
| Aider Polyglot | — | 72% |
| LMArena WebDev | 1397 | — |
| SciCode | 49.9% | — |
| GSO | — | 3.6% |
| CadEval | — | 62% |
| AlgoTune | — | 1.72 |
Agentic & Tool Use o4-mini leads
GPT-5.4 mini: 29.9 (#81), o4-mini: 32.6 (#61)
| Benchmark | GPT-5.4 mini | o4-mini |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 53.2% |
| GDPval | — | 25.3% |
| DeepResearch Bench | 36.3% | — |
| METR Time Horizons | — | 63.9% |
Reasoning GPT-5.4 mini leads
GPT-5.4 mini: 30.4 (#85), o4-mini: 24.6 (#162)
| Benchmark | GPT-5.4 mini | o4-mini |
|---|---|---|
| ARC-AGI-2 | 18.9% | 6.1% |
| Kagi LLM Benchmark | 37.9% | 67.6% |
| ARC-AGI-1 | 63.7% | 58.7% |
| CritPt | 10% | 0.6% |
| Chess Puzzles | 24% | 26% |
| LMArena Hard Prompts | 1424 | 1351 |
| Mystery Game Puzzles | 11% | 5% |
| DTBench | 80% | 77.6% |
| LMCA | 40.8% | 26.5% |
| Epoch Capabilities Index | 148.84 | 145.64 |
| ForecastBench | 57 | 61.8 |
| SimpleBench | — | 38.7% |
| NYT Connections (extended) | 61.8% | — |
| EnigmaEval | — | 9.2% |
| Thematic Generalization | 61.7% | — |
Math GPT-5.4 mini leads
GPT-5.4 mini: 45.5 (#75), o4-mini: 40.8 (#89)
| Benchmark | GPT-5.4 mini | o4-mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 51.2% | 36.1% |
| FrontierMath Tier 4 | 9.8% | 4.9% |
| OTIS Mock AIME 2024-2025 | 88.9% | 81.7% |
| LMArena Math | 1419 | 1389 |
| FrontierMath (Feb 2025 set) | 28.3% | 24.8% |
| FrontierMath Tier 4 (v1) | 2.1% | 6.3% |
| ProofBench | 21% | — |
| Omni-MATH | — | 72% |
| MATH Level 5 | — | 97.8% |
Knowledge GPT-5.4 mini leads
GPT-5.4 mini: 51.5 (#67), o4-mini: 43.6 (#91)
| Benchmark | GPT-5.4 mini | o4-mini |
|---|---|---|
| GPQA Diamond | 86.9% | 79.6% |
| SimpleQA Verified | 29.4% | 19.6% |
| Vectara Hallucination Rate | 5.5% | 18.6% |
| LMArena Expert | 1435 | 1343 |
| Humanity's Last Exam | — | 18.1% |
| MMLU-Pro | — | 82% |
| Confabulations | — | 15.8% |
| GPQA (HELM) | — | 73.5% |
Multimodal Too close to call
GPT-5.4 mini: 39.7 (#56), o4-mini: 40.2 (#49)
| Benchmark | GPT-5.4 mini | o4-mini |
|---|---|---|
| LMArena Vision | 1245 | 1194 |
| GeoBench | — | 64% |
| VPCT | — | 57.5% |
Multilingual GPT-5.4 mini leads
GPT-5.4 mini: 51.9 (#96), o4-mini: 47.0 (#154)
| Benchmark | GPT-5.4 mini | o4-mini |
|---|---|---|
| LMArena Non-English | 1405 | 1337 |
| LMArena Chinese | 1446 | 1354 |
| LMArena French | 1440 | 1364 |
| LMArena German | 1409 | 1336 |
| LMArena Japanese | 1374 | 1308 |
| LMArena Korean | 1368 | 1312 |
| LMArena Russian | 1417 | 1334 |
| LMArena Spanish | 1405 | 1347 |
Instruction Following o4-mini leads
GPT-5.4 mini: 74.1 (#102), o4-mini: 75.2 (#68)
| Benchmark | GPT-5.4 mini | o4-mini |
|---|---|---|
| LMArena Instruction Following | 1405 | 1321 |
| IFEval | — | 92.8% |
Long Context o4-mini leads
GPT-5.4 mini: 43.0 (#112), o4-mini: 45.5 (#33)
| Benchmark | GPT-5.4 mini | o4-mini |
|---|---|---|
| LMArena Longer Query | 1407 | 1315 |
| Fiction.LiveBench | — | 77.8% |
Writing & Preference GPT-5.4 mini leads
GPT-5.4 mini: 64.0 (#58), o4-mini: 54.0 (#152)
| Benchmark | GPT-5.4 mini | o4-mini |
|---|---|---|
| LMArena Text | 1412 | 1353 |
| LMArena Creative Writing | 1370 | 1294 |
| LMArena Multi-Turn | 1429 | 1350 |
| Short-Story Creative Writing | — | 75% |
| EQ-Bench Creative Writing | 1665 | — |
| WildBench | — | 85.4% |
Frequently asked questions
Is GPT-5.4 mini better than o4-mini?
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 41.6 on the Noometry Index.
Which is cheaper, GPT-5.4 mini or o4-mini?
GPT-5.4 mini is cheaper. It lists at $0.75 per million input tokens and $4.50 per million output tokens; o4-mini lists at $1.10 and $4.40.
Is GPT-5.4 mini or o4-mini better for coding?
GPT-5.4 mini scores higher on coding benchmarks: 45.2 versus 40.9 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 mini does, with 400K tokens against 200K.
How many benchmarks do GPT-5.4 mini and o4-mini share?
38 benchmarks have published results for both models. GPT-5.4 mini has 46 scored results on Noometry and o4-mini has 60.