Model comparison
GPT-5.2 vs Llama 3.1-405B
GPT-5.2 is the stronger model overall, scoring 54.1 to 30.4 on the Noometry Index.
Last verified . 26 shared benchmarks.
MetaSummary
- They share 26 benchmarks with published results for both. GPT-5.2 scores higher in 9 categories and Llama 3.1-405B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.2 leads 59.8 to 17.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 96.1% for GPT-5.2 and 9.7% for Llama 3.1-405B.
- Llama 3.1-405B has downloadable open weights; the other is API-only.
Side by side
| GPT-5.2 | Llama 3.1-405B | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 54.1 | 30.4 |
| Released | 2025-12-11 | 2024-07-23 |
| Weights | Proprietary | Open |
| Context window | 400K | — |
| Max output | 128K | — |
| Input $ / M tokens | $1.75 | — |
| Output $ / M tokens | $14 | — |
| Results tracked | 67 | 42 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.2 leads
GPT-5.2: 52.2 (#32), Llama 3.1-405B: 32.7 (#257)
| Benchmark | GPT-5.2 | Llama 3.1-405B |
|---|---|---|
| WeirdML | 72.2% | 21.4% |
| LMArena Coding | 1447 | 1291 |
| SWE-bench Verified | 73.8% | — |
| SWE-bench Verified (bash only) | 72.8% | — |
| LMArena WebDev | 1415 | — |
| SWE-bench Multilingual | 66.7% | — |
| GSO | 27.4% | — |
| ALE-Bench | 1,294 | — |
| AlgoTune | 2.05 | — |
Agentic & Tool Use GPT-5.2 leads
GPT-5.2: 40.0 (#24), Llama 3.1-405B: 20.9 (#131)
| Benchmark | GPT-5.2 | Llama 3.1-405B |
|---|---|---|
| Terminal-Bench | 64.9% | — |
| Berkeley Function Calling Leaderboard | 55.9% | — |
| GDPval | 49.7% | — |
| Remote Labor Index | 2.5% | — |
| TheAgentCompany | — | 7.4% |
| τ²-bench Airline | 83% | — |
| τ²-bench Banking | 32.2% | — |
| τ²-bench Retail | 81.6% | — |
| τ²-bench Telecom | 89.7% | — |
| Cybench | — | 7.5% |
| DeepResearch Bench | 41.1% | — |
| LMArena Search | 1207 | — |
| METR Time Horizons | 75.3% | — |
| Vending-Bench 2 | 3,591 | — |
Reasoning GPT-5.2 leads
GPT-5.2: 50.1 (#33), Llama 3.1-405B: 16.4 (#293)
| Benchmark | GPT-5.2 | Llama 3.1-405B |
|---|---|---|
| SimpleBench | 45.8% | 23% |
| Kagi LLM Benchmark | 73.3% | 45% |
| LMArena Hard Prompts | 1445 | 1270 |
| DTBench | 90.9% | 61.4% |
| Epoch Capabilities Index | 153.45 | 128.76 |
| ForecastBench | 60.1 | 59.9 |
| ARC-AGI-2 | 52.9% | — |
| NYT Connections (extended) | 83.6% | — |
| ARC-AGI-1 | 86.2% | — |
| Chess Puzzles | 49% | — |
| EnigmaEval | 10.4% | — |
| EBR-Bench | 23% | — |
| Mystery Game Puzzles | 23% | — |
| LMCA | 43.9% | — |
| BIG-Bench Hard | — | 82.9% |
| HellaSwag | — | 89.2% |
| PIQA | — | 85.9% |
| WinoGrande | — | 89.2% |
Math GPT-5.2 leads
GPT-5.2: 59.8 (#32), Llama 3.1-405B: 17.3 (#283)
| Benchmark | GPT-5.2 | Llama 3.1-405B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 96.1% | 9.7% |
| LMArena Math | 1439 | 1281 |
| FrontierMath (Tiers 1-3) | 67.4% | — |
| FrontierMath Tier 4 | 31.7% | — |
| MathArena Final-Answer Competitions | 72% | — |
| ProofBench | 15% | — |
| Omni-MATH | — | 24.9% |
| MATH Level 5 | — | 49.8% |
| FrontierMath (Feb 2025 set) | 40.7% | — |
| FrontierMath Tier 4 (v1) | 18.8% | — |
Knowledge GPT-5.2 leads
GPT-5.2: 59.2 (#29), Llama 3.1-405B: 30.4 (#221)
| Benchmark | GPT-5.2 | Llama 3.1-405B |
|---|---|---|
| GPQA Diamond | 91.4% | 50.9% |
| LMArena Expert | 1445 | 1243 |
| Humanity's Last Exam | 27.8% | — |
| SimpleQA Verified | 37.1% | — |
| MMLU-Pro | — | 72.3% |
| Confabulations | — | 17.6% |
| Vectara Hallucination Rate | 8.4% | — |
| GPQA (HELM) | — | 52.2% |
| ARC (AI2) Challenge | — | 95.3% |
| MMLU | — | 84.5% |
| TriviaQA | — | 82.7% |
Multimodal Not comparable
GPT-5.2: 50.9 (#7), Llama 3.1-405B: —
| Benchmark | GPT-5.2 | Llama 3.1-405B |
|---|---|---|
| LMArena Vision | 1268 | — |
| VPCT | 84% | — |
| Furniture Assembly | 38.3% | — |
| LMArena Document | 1405 | — |
Multilingual GPT-5.2 leads
GPT-5.2: 53.6 (#63), Llama 3.1-405B: 40.9 (#209)
| Benchmark | GPT-5.2 | Llama 3.1-405B |
|---|---|---|
| LMArena Non-English | 1425 | 1248 |
| LMArena Chinese | 1460 | 1242 |
| LMArena French | 1456 | 1278 |
| LMArena German | 1448 | 1252 |
| LMArena Japanese | 1421 | 1208 |
| LMArena Korean | 1390 | 1184 |
| LMArena Russian | 1440 | 1265 |
| LMArena Spanish | 1433 | 1259 |
Instruction Following GPT-5.2 leads
GPT-5.2: 74.9 (#83), Llama 3.1-405B: 66.1 (#209)
| Benchmark | GPT-5.2 | Llama 3.1-405B |
|---|---|---|
| LMArena Instruction Following | 1417 | 1259 |
| IFEval | — | 81.1% |
Long Context GPT-5.2 leads
GPT-5.2: 44.0 (#74), Llama 3.1-405B: 38.4 (#192)
| Benchmark | GPT-5.2 | Llama 3.1-405B |
|---|---|---|
| LMArena Longer Query | 1428 | 1266 |
| CL-bench | 18.2% | — |
Writing & Preference GPT-5.2 leads
GPT-5.2: 67.0 (#30), Llama 3.1-405B: 39.1 (#246)
| Benchmark | GPT-5.2 | Llama 3.1-405B |
|---|---|---|
| LMArena Text | 1439 | 1284 |
| LMArena Creative Writing | 1402 | 1262 |
| EQ-Bench Creative Writing | 1703 | 870 |
| LMArena Multi-Turn | 1458 | 1297 |
| WildBench | — | 78.3% |
Frequently asked questions
Is GPT-5.2 better than Llama 3.1-405B?
GPT-5.2 is the stronger model overall, scoring 54.1 to 30.4 on the Noometry Index.
Is GPT-5.2 or Llama 3.1-405B better for coding?
GPT-5.2 scores higher on coding benchmarks: 52.2 versus 32.7 in the Noometry coding category.
How many benchmarks do GPT-5.2 and Llama 3.1-405B share?
26 benchmarks have published results for both models. GPT-5.2 has 67 scored results on Noometry and Llama 3.1-405B has 42.