Model comparison
Llama 3.1-405B vs o3-pro
o3-pro is the stronger model overall, scoring 42.9 to 30.7 on the Noometry Index.
Last verified . 5 shared benchmarks.
MetaSummary
- They share 5 benchmarks with published results for both. Llama 3.1-405B scores higher in 1 category and o3-pro in 4 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in long context, where o3-pro leads 72.2 to 38.4.
- The biggest single-benchmark swing is WeirdML: 21.4% for Llama 3.1-405B and 58.2% for o3-pro.
- Llama 3.1-405B has downloadable open weights; the other is API-only.
Side by side
| Llama 3.1-405B | o3-pro | |
|---|---|---|
| Provider | Meta | OpenAI |
| Noometry Index | 30.7 | 42.9 |
| Released | 2024-07-23 | 2025-06-10 |
| Weights | Open | Proprietary |
| Context window | — | 200K |
| Max output | — | 100K |
| Input $ / M tokens | — | $20 |
| Output $ / M tokens | — | $80 |
| Results tracked | 42 | 12 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o3-pro leads
Llama 3.1-405B: 33.1 (#262), o3-pro: 55.5 (#24)
| Benchmark | Llama 3.1-405B | o3-pro |
|---|---|---|
| WeirdML | 21.4% | 58.2% |
| Aider Polyglot | — | 84.9% |
| LMArena Coding | 1291 | — |
Agentic & Tool Use Not comparable
Llama 3.1-405B: 21.0 (#140), o3-pro: —
| Benchmark | Llama 3.1-405B | o3-pro |
|---|---|---|
| TheAgentCompany | 7.4% | — |
| Cybench | 7.5% | — |
Reasoning o3-pro leads
Llama 3.1-405B: 16.8 (#300), o3-pro: 23.8 (#171)
| Benchmark | Llama 3.1-405B | o3-pro |
|---|---|---|
| Kagi LLM Benchmark | 45% | 72.1% |
| DTBench | 61.4% | 86.9% |
| Epoch Capabilities Index | 128.75 | 147.42 |
| ARC-AGI-2 | — | 4.9% |
| SimpleBench | 23% | — |
| ARC-AGI-1 | — | 59.3% |
| LMArena Hard Prompts | 1269 | — |
| LMCA | — | 38.5% |
| BIG-Bench Hard | 82.9% | — |
| ForecastBench | 59.9 | — |
| HellaSwag | 89.2% | — |
| PIQA | 85.9% | — |
| WinoGrande | 89.2% | — |
Math Not comparable
Llama 3.1-405B: 18.4 (#290), o3-pro: —
| Benchmark | Llama 3.1-405B | o3-pro |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 9.7% | — |
| Omni-MATH | 24.9% | — |
| LMArena Math | 1281 | — |
| MATH Level 5 | 49.8% | — |
Knowledge Too close to call
Llama 3.1-405B: 30.4 (#227), o3-pro: 29.5 (#238)
| Benchmark | Llama 3.1-405B | o3-pro |
|---|---|---|
| Confabulations | 17.6% | 14.2% |
| GPQA Diamond | 50.9% | — |
| MMLU-Pro | 72.3% | — |
| Vectara Hallucination Rate | — | 23.3% |
| GPQA (HELM) | 52.2% | — |
| LMArena Expert | 1243 | — |
| ARC (AI2) Challenge | 95.3% | — |
| MMLU | 84.5% | — |
| TriviaQA | 82.7% | — |
Multilingual Not comparable
Llama 3.1-405B: 40.7 (#214), o3-pro: —
| Benchmark | Llama 3.1-405B | o3-pro |
|---|---|---|
| LMArena Non-English | 1248 | — |
| LMArena Chinese | 1242 | — |
| LMArena French | 1279 | — |
| LMArena German | 1252 | — |
| LMArena Japanese | 1208 | — |
| LMArena Korean | 1184 | — |
| LMArena Russian | 1265 | — |
| LMArena Spanish | 1260 | — |
Instruction Following Not comparable
Llama 3.1-405B: 65.9 (#214), o3-pro: —
| Benchmark | Llama 3.1-405B | o3-pro |
|---|---|---|
| IFEval | 81.1% | — |
| LMArena Instruction Following | 1259 | — |
Long Context o3-pro leads
Llama 3.1-405B: 38.4 (#197), o3-pro: 72.2 (#1)
| Benchmark | Llama 3.1-405B | o3-pro |
|---|---|---|
| Fiction.LiveBench | — | 97.2% |
| LMArena Longer Query | 1266 | — |
Writing & Preference o3-pro leads
Llama 3.1-405B: 38.9 (#251), o3-pro: 57.1 (#133)
| Benchmark | Llama 3.1-405B | o3-pro |
|---|---|---|
| LMArena Text | 1284 | — |
| LMArena Creative Writing | 1262 | — |
| Short-Story Creative Writing | — | 84.4% |
| EQ-Bench Creative Writing | 870 | — |
| WildBench | 78.3% | — |
| LMArena Multi-Turn | 1297 | — |
Frequently asked questions
Is Llama 3.1-405B better than o3-pro?
o3-pro is the stronger model overall, scoring 42.9 to 30.7 on the Noometry Index.
Is Llama 3.1-405B or o3-pro better for coding?
o3-pro scores higher on coding benchmarks: 55.5 versus 33.1 in the Noometry coding category.
How many benchmarks do Llama 3.1-405B and o3-pro share?
5 benchmarks have published results for both models. Llama 3.1-405B has 42 scored results on Noometry and o3-pro has 12.