Model comparison
GPT-5 Nano vs Llama 4 Maverick
GPT-5 Nano is the stronger model overall, scoring 33.4 to 30.6 on the Noometry Index.
Last verified . 39 shared benchmarks.
MetaSummary
- They share 39 benchmarks with published results for both. GPT-5 Nano scores higher in 7 categories and Llama 4 Maverick in 3 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in coding, where GPT-5 Nano leads 33.3 to 26.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 81.1% for GPT-5 Nano and 20.6% for Llama 4 Maverick.
- GPT-5 Nano is cheaper at $0.05 / $0.40 per million input/output tokens, against $0.19 / $0.65 for Llama 4 Maverick.
- GPT-5 Nano accepts more context: 400K tokens versus 128K.
- Llama 4 Maverick has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Nano | Llama 4 Maverick | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 33.4 | 30.6 |
| Released | 2025-08-07 | 2025-04-05 |
| Weights | Proprietary | Open |
| Context window | 400K | 128K |
| Max output | 128K | 4K |
| Input $ / M tokens | $0.05 | $0.19 |
| Output $ / M tokens | $0.40 | $0.65 |
| Results tracked | 48 | 54 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 Nano leads
GPT-5 Nano: 33.3 (#248), Llama 4 Maverick: 26.3 (#318)
| Benchmark | GPT-5 Nano | Llama 4 Maverick |
|---|---|---|
| SWE-bench Verified (bash only) | 34.8% | 21% |
| WeirdML | 38.1% | 24.5% |
| LMArena Coding | 1351 | 1302 |
| ALE-Bench | 718.67 | 172.97 |
| Aider Polyglot | — | 15.6% |
| SciCode | — | 33.1% |
| BigCodeBench Instruct | — | 49.7% |
| BigCodeBench Complete | — | 61.4% |
Agentic & Tool Use Llama 4 Maverick leads
GPT-5 Nano: 25.6 (#99), Llama 4 Maverick: 27.9 (#86)
| Benchmark | GPT-5 Nano | Llama 4 Maverick |
|---|---|---|
| Berkeley Function Calling Leaderboard | 51.5% | 37.3% |
| Terminal-Bench | 21.8% | — |
Reasoning GPT-5 Nano leads
GPT-5 Nano: 16.4 (#291), Llama 4 Maverick: 10.1 (#337)
| Benchmark | GPT-5 Nano | Llama 4 Maverick |
|---|---|---|
| ARC-AGI-2 | 2.6% | 0% |
| Kagi LLM Benchmark | 62.2% | 55.9% |
| ARC-AGI-1 | 20.7% | 4.4% |
| LMArena Hard Prompts | 1329 | 1281 |
| DTBench | 62.7% | 61.9% |
| LMCA | 7.9% | 15.9% |
| Epoch Capabilities Index | 139.39 | 132.2 |
| ForecastBench | 59.1 | 57.5 |
| SimpleBench | — | 27.7% |
| NYT Connections (extended) | — | 8% |
| CritPt | — | 0% |
| Chess Puzzles | 27% | — |
| EnigmaEval | — | 0.6% |
| Mystery Game Puzzles | 9% | — |
Math GPT-5 Nano leads
GPT-5 Nano: 29.1 (#221), Llama 4 Maverick: 24.6 (#256)
| Benchmark | GPT-5 Nano | Llama 4 Maverick |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 81.1% | 20.6% |
| Omni-MATH | 54.6% | 42.2% |
| LMArena Math | 1318 | 1299 |
| MATH Level 5 | 95.2% | 73% |
| FrontierMath (Feb 2025 set) | 8.3% | 0.7% |
| FrontierMath (Tiers 1-3) | 20% | — |
| FrontierMath Tier 4 | 2.4% | — |
| ProofBench | 12% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge GPT-5 Nano leads
GPT-5 Nano: 35.8 (#166), Llama 4 Maverick: 33.3 (#196)
| Benchmark | GPT-5 Nano | Llama 4 Maverick |
|---|---|---|
| GPQA Diamond | 69.4% | 67% |
| MMLU-Pro | 77.8% | 81% |
| Vectara Hallucination Rate | 10.5% | 8.2% |
| GPQA (HELM) | 67.9% | 65% |
| LMArena Expert | 1321 | 1259 |
| Humanity's Last Exam | — | 5.7% |
| SimpleQA Verified | 11.7% | — |
| Confabulations | — | 22.6% |
Multimodal Too close to call
GPT-5 Nano: 30.7 (#102), Llama 4 Maverick: 30.9 (#100)
| Benchmark | GPT-5 Nano | Llama 4 Maverick |
|---|---|---|
| LMArena Vision | 1159 | 1141 |
| GeoBench | — | 52% |
| VPCT | 37.2% | — |
| SpatialViz-Bench | — | 31.8% |
Multilingual GPT-5 Nano leads
GPT-5 Nano: 45.5 (#167), Llama 4 Maverick: 42.4 (#190)
| Benchmark | GPT-5 Nano | Llama 4 Maverick |
|---|---|---|
| LMArena Non-English | 1313 | 1269 |
| LMArena Chinese | 1355 | 1277 |
| LMArena German | 1329 | 1291 |
| LMArena Korean | 1269 | 1203 |
| LMArena Russian | 1296 | 1286 |
| LMArena Spanish | 1360 | 1293 |
| LMArena French | — | 1258 |
| LMArena Japanese | — | 1207 |
Instruction Following GPT-5 Nano leads
GPT-5 Nano: 75.2 (#72), Llama 4 Maverick: 71.9 (#140)
| Benchmark | GPT-5 Nano | Llama 4 Maverick |
|---|---|---|
| IFEval | 93.2% | 90.8% |
| LMArena Instruction Following | 1306 | 1267 |
Long Context Too close to call
GPT-5 Nano: 31.3 (#275), Llama 4 Maverick: 31.4 (#274)
| Benchmark | GPT-5 Nano | Llama 4 Maverick |
|---|---|---|
| Fiction.LiveBench | 44.4% | 46.2% |
| LMArena Longer Query | 1313 | 1280 |
Writing & Preference Too close to call
GPT-5 Nano: 39.4 (#244), Llama 4 Maverick: 39.0 (#247)
| Benchmark | GPT-5 Nano | Llama 4 Maverick |
|---|---|---|
| LMArena Text | 1320 | 1287 |
| LMArena Creative Writing | 1250 | 1266 |
| EQ-Bench Creative Writing | 705 | 860 |
| WildBench | 80.6% | 80% |
| LMArena Multi-Turn | 1311 | 1289 |
| Short-Story Creative Writing | — | 62% |
Frequently asked questions
Is GPT-5 Nano better than Llama 4 Maverick?
GPT-5 Nano is the stronger model overall, scoring 33.4 to 30.6 on the Noometry Index.
Which is cheaper, GPT-5 Nano or Llama 4 Maverick?
GPT-5 Nano is cheaper. It lists at $0.05 per million input tokens and $0.40 per million output tokens; Llama 4 Maverick lists at $0.19 and $0.65.
Is GPT-5 Nano or Llama 4 Maverick better for coding?
GPT-5 Nano scores higher on coding benchmarks: 33.3 versus 26.3 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Nano does, with 400K tokens against 128K.
How many benchmarks do GPT-5 Nano and Llama 4 Maverick share?
39 benchmarks have published results for both models. GPT-5 Nano has 48 scored results on Noometry and Llama 4 Maverick has 54.