Model comparison
GPT-4.5 vs GPT-5.4 mini
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 37.2 on the Noometry Index.
Last verified . 25 shared benchmarks.
Summary
- They share 25 benchmarks with published results for both. GPT-4.5 scores higher in 1 category and GPT-5.4 mini in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-5.4 mini leads 51.5 to 32.5.
- The biggest single-benchmark swing is ARC-AGI-1: 10.3% for GPT-4.5 and 63.7% for GPT-5.4 mini.
Side by side
| GPT-4.5 | GPT-5.4 mini | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 37.2 | 45.0 |
| Released | 2025-02-27 | 2026-03-17 |
| Weights | Proprietary | Proprietary |
| Context window | — | 400K |
| Max output | — | 128K |
| Input $ / M tokens | — | $0.75 |
| Output $ / M tokens | — | $4.50 |
| Results tracked | 42 | 46 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 mini leads
GPT-4.5: 42.2 (#109), GPT-5.4 mini: 45.2 (#72)
| Benchmark | GPT-4.5 | GPT-5.4 mini |
|---|---|---|
| WeirdML | 39.4% | 60.3% |
| LMArena Coding | 1396 | 1438 |
| FrontierCode | — | 27% |
| Aider Polyglot | 44.9% | — |
| LMArena WebDev | — | 1397 |
| SciCode | — | 49.9% |
| LiveBench Coding | 75.2% | — |
| ALE-Bench | — | 1,189 |
Agentic & Tool Use GPT-5.4 mini leads
GPT-4.5: 27.9 (#97), GPT-5.4 mini: 29.9 (#81)
| Benchmark | GPT-4.5 | GPT-5.4 mini |
|---|---|---|
| Cybench | 17.5% | — |
| DeepResearch Bench | — | 36.3% |
Reasoning GPT-5.4 mini leads
GPT-4.5: 13.9 (#330), GPT-5.4 mini: 30.4 (#85)
| Benchmark | GPT-4.5 | GPT-5.4 mini |
|---|---|---|
| ARC-AGI-2 | 0.8% | 18.9% |
| ARC-AGI-1 | 10.3% | 63.7% |
| LMArena Hard Prompts | 1403 | 1424 |
| Epoch Capabilities Index | 136.74 | 148.84 |
| ForecastBench | 61.7 | 57 |
| SimpleBench | 34.5% | — |
| Kagi LLM Benchmark | — | 37.9% |
| NYT Connections (extended) | — | 61.8% |
| CritPt | — | 10% |
| Chess Puzzles | — | 24% |
| EnigmaEval | 3.2% | — |
| Thematic Generalization | — | 61.7% |
| LiveBench Reasoning | 71.1% | — |
| Mystery Game Puzzles | — | 11% |
| DTBench | — | 80% |
| LiveBench Data Analysis | 64.3% | — |
| LMCA | — | 40.8% |
| LiveBench | 69% | — |
Math GPT-5.4 mini leads
GPT-4.5: 32.6 (#211), GPT-5.4 mini: 45.5 (#75)
| Benchmark | GPT-4.5 | GPT-5.4 mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 88.9% |
| LMArena Math | 1412 | 1419 |
| FrontierMath (Tiers 1-3) | — | 51.2% |
| FrontierMath Tier 4 | — | 9.8% |
| ProofBench | — | 21% |
| LiveBench Math | 69.3% | — |
| MATH Level 5 | 78.6% | — |
| FrontierMath (Feb 2025 set) | — | 28.3% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge GPT-5.4 mini leads
GPT-4.5: 32.5 (#211), GPT-5.4 mini: 51.5 (#67)
| Benchmark | GPT-4.5 | GPT-5.4 mini |
|---|---|---|
| GPQA Diamond | 68.7% | 86.9% |
| LMArena Expert | 1394 | 1435 |
| Humanity's Last Exam | 5.4% | — |
| SimpleQA Verified | — | 29.4% |
| Confabulations | 13.6% | — |
| Vectara Hallucination Rate | — | 5.5% |
Multimodal GPT-5.4 mini leads
GPT-4.5: 37.6 (#71), GPT-5.4 mini: 39.7 (#56)
| Benchmark | GPT-4.5 | GPT-5.4 mini |
|---|---|---|
| LMArena Vision | 1195 | 1245 |
| VPCT | 45% | — |
Multilingual Too close to call
GPT-4.5: 52.5 (#83), GPT-5.4 mini: 51.9 (#96)
| Benchmark | GPT-4.5 | GPT-5.4 mini |
|---|---|---|
| LMArena Non-English | 1413 | 1405 |
| LMArena Chinese | 1421 | 1446 |
| LMArena French | 1418 | 1440 |
| LMArena German | 1457 | 1409 |
| LMArena Japanese | 1416 | 1374 |
| LMArena Korean | 1392 | 1368 |
| LMArena Russian | 1419 | 1417 |
| LMArena Spanish | — | 1405 |
Instruction Following GPT-5.4 mini leads
GPT-4.5: 72.6 (#134), GPT-5.4 mini: 74.1 (#102)
| Benchmark | GPT-4.5 | GPT-5.4 mini |
|---|---|---|
| LMArena Instruction Following | 1404 | 1405 |
| LiveBench Instruction Following | 72.3% | — |
Long Context GPT-5.4 mini leads
GPT-4.5: 40.4 (#155), GPT-5.4 mini: 43.0 (#112)
| Benchmark | GPT-4.5 | GPT-5.4 mini |
|---|---|---|
| LMArena Longer Query | 1406 | 1407 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference GPT-5.4 mini leads
GPT-4.5: 56.9 (#134), GPT-5.4 mini: 64.0 (#58)
| Benchmark | GPT-4.5 | GPT-5.4 mini |
|---|---|---|
| LMArena Text | 1417 | 1412 |
| LMArena Creative Writing | 1394 | 1370 |
| EQ-Bench Creative Writing | 1258 | 1665 |
| LMArena Multi-Turn | 1444 | 1429 |
| Short-Story Creative Writing | 75.6% | — |
| LiveBench Language | 61.5% | — |
Frequently asked questions
Is GPT-4.5 better than GPT-5.4 mini?
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 37.2 on the Noometry Index.
Is GPT-4.5 or GPT-5.4 mini better for coding?
GPT-5.4 mini scores higher on coding benchmarks: 45.2 versus 42.2 in the Noometry coding category.
How many benchmarks do GPT-4.5 and GPT-5.4 mini share?
25 benchmarks have published results for both models. GPT-4.5 has 42 scored results on Noometry and GPT-5.4 mini has 46.