Model comparison
GPT-4.5 vs Llama 4 Maverick
GPT-4.5 is the stronger model overall, scoring 37.2 to 30.9 on the Noometry Index.
Last verified . 33 shared benchmarks.
MetaSummary
- They share 33 benchmarks with published results for both. GPT-4.5 scores higher in 8 categories and Llama 4 Maverick in 2 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GPT-4.5 leads 56.9 to 38.8.
- The biggest single-benchmark swing is Aider Polyglot: 44.9% for GPT-4.5 and 15.6% for Llama 4 Maverick.
- Llama 4 Maverick has downloadable open weights; the other is API-only.
Side by side
| GPT-4.5 | Llama 4 Maverick | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 37.2 | 30.9 |
| Released | 2025-02-27 | 2025-04-05 |
| Weights | Proprietary | Open |
| Context window | — | 128K |
| Max output | — | 4K |
| Input $ / M tokens | — | $0.19 |
| Output $ / M tokens | — | $0.65 |
| Results tracked | 42 | 54 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-4.5 leads
GPT-4.5: 42.2 (#109), Llama 4 Maverick: 26.6 (#324)
| Benchmark | GPT-4.5 | Llama 4 Maverick |
|---|---|---|
| Aider Polyglot | 44.9% | 15.6% |
| WeirdML | 39.4% | 24.5% |
| LMArena Coding | 1396 | 1302 |
| SWE-bench Verified (bash only) | — | 21% |
| SciCode | — | 33.1% |
| BigCodeBench Instruct | — | 49.7% |
| LiveBench Coding | 75.2% | — |
| BigCodeBench Complete | — | 61.4% |
| ALE-Bench | — | 172.97 |
Agentic & Tool Use Too close to call
GPT-4.5: 27.9 (#97), Llama 4 Maverick: 28.2 (#91)
| Benchmark | GPT-4.5 | Llama 4 Maverick |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 37.3% |
| Cybench | 17.5% | — |
Reasoning GPT-4.5 leads
GPT-4.5: 13.9 (#330), Llama 4 Maverick: 10.1 (#342)
| Benchmark | GPT-4.5 | Llama 4 Maverick |
|---|---|---|
| ARC-AGI-2 | 0.8% | 0% |
| SimpleBench | 34.5% | 27.7% |
| ARC-AGI-1 | 10.3% | 4.4% |
| EnigmaEval | 3.2% | 0.6% |
| LMArena Hard Prompts | 1403 | 1281 |
| Epoch Capabilities Index | 136.74 | 132.2 |
| ForecastBench | 61.7 | 57.5 |
| Kagi LLM Benchmark | — | 55.9% |
| NYT Connections (extended) | — | 8% |
| CritPt | — | 0% |
| LiveBench Reasoning | 71.1% | — |
| DTBench | — | 61.9% |
| LiveBench Data Analysis | 64.3% | — |
| LMCA | — | 15.9% |
| LiveBench | 69% | — |
Math GPT-4.5 leads
GPT-4.5: 32.6 (#211), Llama 4 Maverick: 26.0 (#262)
| Benchmark | GPT-4.5 | Llama 4 Maverick |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 20.6% |
| LMArena Math | 1412 | 1299 |
| MATH Level 5 | 78.6% | 73% |
| Omni-MATH | — | 42.2% |
| LiveBench Math | 69.3% | — |
| FrontierMath (Feb 2025 set) | — | 0.7% |
Knowledge Too close to call
GPT-4.5: 32.5 (#211), Llama 4 Maverick: 33.4 (#204)
| Benchmark | GPT-4.5 | Llama 4 Maverick |
|---|---|---|
| GPQA Diamond | 68.7% | 67% |
| Humanity's Last Exam | 5.4% | 5.7% |
| Confabulations | 13.6% | 22.6% |
| LMArena Expert | 1394 | 1259 |
| MMLU-Pro | — | 81% |
| Vectara Hallucination Rate | — | 8.2% |
| GPQA (HELM) | — | 65% |
Multimodal GPT-4.5 leads
GPT-4.5: 37.6 (#71), Llama 4 Maverick: 31.6 (#105)
| Benchmark | GPT-4.5 | Llama 4 Maverick |
|---|---|---|
| LMArena Vision | 1195 | 1142 |
| GeoBench | — | 52% |
| VPCT | 45% | — |
| SpatialViz-Bench | — | 31.8% |
Multilingual GPT-4.5 leads
GPT-4.5: 52.5 (#83), Llama 4 Maverick: 42.2 (#195)
| Benchmark | GPT-4.5 | Llama 4 Maverick |
|---|---|---|
| LMArena Non-English | 1413 | 1269 |
| LMArena Chinese | 1421 | 1277 |
| LMArena French | 1418 | 1259 |
| LMArena German | 1457 | 1291 |
| LMArena Japanese | 1416 | 1207 |
| LMArena Korean | 1392 | 1203 |
| LMArena Russian | 1419 | 1286 |
| LMArena Spanish | — | 1293 |
Instruction Following Too close to call
GPT-4.5: 72.6 (#134), Llama 4 Maverick: 71.7 (#146)
| Benchmark | GPT-4.5 | Llama 4 Maverick |
|---|---|---|
| LMArena Instruction Following | 1404 | 1267 |
| LiveBench Instruction Following | 72.3% | — |
| IFEval | — | 90.8% |
Long Context GPT-4.5 leads
GPT-4.5: 40.4 (#155), Llama 4 Maverick: 31.4 (#279)
| Benchmark | GPT-4.5 | Llama 4 Maverick |
|---|---|---|
| Fiction.LiveBench | 63.9% | 46.2% |
| LMArena Longer Query | 1406 | 1280 |
Writing & Preference GPT-4.5 leads
GPT-4.5: 56.9 (#134), Llama 4 Maverick: 38.8 (#252)
| Benchmark | GPT-4.5 | Llama 4 Maverick |
|---|---|---|
| LMArena Text | 1417 | 1287 |
| LMArena Creative Writing | 1394 | 1267 |
| Short-Story Creative Writing | 75.6% | 62% |
| EQ-Bench Creative Writing | 1258 | 860 |
| LMArena Multi-Turn | 1444 | 1289 |
| WildBench | — | 80% |
| LiveBench Language | 61.5% | — |
Frequently asked questions
Is GPT-4.5 better than Llama 4 Maverick?
GPT-4.5 is the stronger model overall, scoring 37.2 to 30.9 on the Noometry Index.
Is GPT-4.5 or Llama 4 Maverick better for coding?
GPT-4.5 scores higher on coding benchmarks: 42.2 versus 26.6 in the Noometry coding category.
How many benchmarks do GPT-4.5 and Llama 4 Maverick share?
33 benchmarks have published results for both models. GPT-4.5 has 42 scored results on Noometry and Llama 4 Maverick has 54.