Model comparison
GPT-4 Turbo vs Llama 3.1-405B
GPT-4 Turbo and Llama 3.1-405B score almost the same on the Noometry Index (30.5 vs 30.7), so choose on price, context window or the category you care about most.
Last verified . 27 shared benchmarks.
MetaSummary
- They share 27 benchmarks with published results for both. GPT-4 Turbo scores higher in 2 categories and Llama 3.1-405B in 6 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in math, where Llama 3.1-405B leads 18.4 to 9.0.
- The biggest single-benchmark swing is Confabulations: 28.4% for GPT-4 Turbo and 17.6% for Llama 3.1-405B.
- Llama 3.1-405B has downloadable open weights; the other is API-only.
Side by side
| GPT-4 Turbo | Llama 3.1-405B | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 30.5 | 30.7 |
| Released | 2023-11-06 | 2024-07-23 |
| Weights | Proprietary | Open |
| Context window | 128K | — |
| Max output | 4K | — |
| Input $ / M tokens | $10 | — |
| Output $ / M tokens | $30 | — |
| Results tracked | 36 | 42 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GPT-4 Turbo: 33.8 (#249), Llama 3.1-405B: 33.1 (#262)
| Benchmark | GPT-4 Turbo | Llama 3.1-405B |
|---|---|---|
| WeirdML | 18% | 21.4% |
| LMArena Coding | 1268 | 1291 |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 58.2% | — |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73.3% | — |
Agentic & Tool Use Not comparable
GPT-4 Turbo: —, Llama 3.1-405B: 21.0 (#140)
| Benchmark | GPT-4 Turbo | Llama 3.1-405B |
|---|---|---|
| TheAgentCompany | — | 7.4% |
| Cybench | — | 7.5% |
| METR Time Horizons | 36.7% | — |
Reasoning Llama 3.1-405B leads
GPT-4 Turbo: 15.3 (#317), Llama 3.1-405B: 16.8 (#300)
| Benchmark | GPT-4 Turbo | Llama 3.1-405B |
|---|---|---|
| SimpleBench | 25.1% | 23% |
| LMArena Hard Prompts | 1251 | 1269 |
| DTBench | 61.6% | 61.4% |
| Epoch Capabilities Index | 127.25 | 128.75 |
| ForecastBench | 59.4 | 59.9 |
| Kagi LLM Benchmark | — | 45% |
| Chess Puzzles | 6% | — |
| LMCA | 9.8% | — |
| BIG-Bench Hard | — | 82.9% |
| HellaSwag | — | 89.2% |
| PIQA | — | 85.9% |
| WinoGrande | — | 89.2% |
Math Llama 3.1-405B leads
GPT-4 Turbo: 9.0 (#322), Llama 3.1-405B: 18.4 (#290)
| Benchmark | GPT-4 Turbo | Llama 3.1-405B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 6.7% | 9.7% |
| LMArena Math | 1272 | 1281 |
| MATH Level 5 | 46.7% | 49.8% |
| FrontierMath (Tiers 1-3) | 0.7% | — |
| Omni-MATH | — | 24.9% |
Knowledge Llama 3.1-405B leads
GPT-4 Turbo: 24.3 (#268), Llama 3.1-405B: 30.4 (#227)
| Benchmark | GPT-4 Turbo | Llama 3.1-405B |
|---|---|---|
| GPQA Diamond | 46.6% | 50.9% |
| Confabulations | 28.4% | 17.6% |
| LMArena Expert | 1223 | 1243 |
| MMLU | 81.3% | 84.5% |
| MMLU-Pro | — | 72.3% |
| GPQA (HELM) | — | 52.2% |
| ARC (AI2) Challenge | — | 95.3% |
| TriviaQA | — | 82.7% |
Multimodal Not comparable
GPT-4 Turbo: 30.6 (#110), Llama 3.1-405B: —
| Benchmark | GPT-4 Turbo | Llama 3.1-405B |
|---|---|---|
| LMArena Vision | 1090 | — |
Multilingual Too close to call
GPT-4 Turbo: 40.5 (#216), Llama 3.1-405B: 40.7 (#214)
| Benchmark | GPT-4 Turbo | Llama 3.1-405B |
|---|---|---|
| LMArena Non-English | 1245 | 1248 |
| LMArena Chinese | 1242 | 1242 |
| LMArena French | 1276 | 1279 |
| LMArena German | 1259 | 1252 |
| LMArena Japanese | 1194 | 1208 |
| LMArena Korean | 1187 | 1184 |
| LMArena Russian | 1259 | 1265 |
| LMArena Spanish | 1260 | 1260 |
Instruction Following Too close to call
GPT-4 Turbo: 65.8 (#216), Llama 3.1-405B: 65.9 (#214)
| Benchmark | GPT-4 Turbo | Llama 3.1-405B |
|---|---|---|
| LMArena Instruction Following | 1249 | 1259 |
| IFEval | — | 81.1% |
Long Context Too close to call
GPT-4 Turbo: 38.0 (#206), Llama 3.1-405B: 38.4 (#197)
| Benchmark | GPT-4 Turbo | Llama 3.1-405B |
|---|---|---|
| LMArena Longer Query | 1254 | 1266 |
Writing & Preference GPT-4 Turbo leads
GPT-4 Turbo: 47.7 (#206), Llama 3.1-405B: 38.9 (#251)
| Benchmark | GPT-4 Turbo | Llama 3.1-405B |
|---|---|---|
| LMArena Text | 1272 | 1284 |
| LMArena Creative Writing | 1269 | 1262 |
| LMArena Multi-Turn | 1267 | 1297 |
| EQ-Bench Creative Writing | — | 870 |
| WildBench | — | 78.3% |
Frequently asked questions
Is GPT-4 Turbo better than Llama 3.1-405B?
GPT-4 Turbo and Llama 3.1-405B score almost the same on the Noometry Index (30.5 vs 30.7), so choose on price, context window or the category you care about most.
Is GPT-4 Turbo or Llama 3.1-405B better for coding?
They score almost the same on coding (33.8 vs 33.1); test both on your own repository before choosing.
How many benchmarks do GPT-4 Turbo and Llama 3.1-405B share?
27 benchmarks have published results for both models. GPT-4 Turbo has 36 scored results on Noometry and Llama 3.1-405B has 42.