Model comparison
Llama 3.1-405B vs Llama 3.1-70B
Llama 3.1-405B is the stronger model overall, scoring 30.7 to 29.6 on the Noometry Index.
Last verified . 31 shared benchmarks.
MetaSummary
- They share 31 benchmarks with published results for both. Llama 3.1-405B scores higher in 7 categories and Llama 3.1-70B in 2 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Llama 3.1-405B leads 30.4 to 24.2.
- The biggest single-benchmark swing is MATH Level 5: 49.8% for Llama 3.1-405B and 36.7% for Llama 3.1-70B.
Side by side
| Llama 3.1-405B | Llama 3.1-70B | |
|---|---|---|
| Provider | Meta | Meta |
| Noometry Index | 30.7 | 29.6 |
| Released | 2024-07-23 | 2024-07-23 |
| Weights | Open | Open |
| Context window | — | 128K |
| Max output | — | 4K |
| Input $ / M tokens | — | $0.40 |
| Output $ / M tokens | — | $0.40 |
| Results tracked | 42 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 3.1-405B leads
Llama 3.1-405B: 33.1 (#262), Llama 3.1-70B: 30.3 (#296)
| Benchmark | Llama 3.1-405B | Llama 3.1-70B |
|---|---|---|
| WeirdML | 21.4% | 9% |
| LMArena Coding | 1291 | 1260 |
| BigCodeBench Instruct | — | 46.1% |
| BigCodeBench Complete | — | 54.8% |
Agentic & Tool Use Llama 3.1-70B leads
Llama 3.1-405B: 21.0 (#140), Llama 3.1-70B: 25.1 (#112)
| Benchmark | Llama 3.1-405B | Llama 3.1-70B |
|---|---|---|
| TheAgentCompany | 7.4% | 6.9% |
| Cybench | 7.5% | — |
| BALROG | — | 27.9% |
Reasoning Llama 3.1-70B leads
Llama 3.1-405B: 16.8 (#300), Llama 3.1-70B: 21.6 (#220)
| Benchmark | Llama 3.1-405B | Llama 3.1-70B |
|---|---|---|
| LMArena Hard Prompts | 1269 | 1241 |
| DTBench | 61.4% | 60% |
| Epoch Capabilities Index | 128.75 | 125.92 |
| SimpleBench | 23% | — |
| Kagi LLM Benchmark | 45% | — |
| LMCA | — | 14.8% |
| BIG-Bench Hard | 82.9% | — |
| ForecastBench | 59.9 | — |
| HellaSwag | 89.2% | — |
| PIQA | 85.9% | — |
| WinoGrande | 89.2% | — |
Math Llama 3.1-405B leads
Llama 3.1-405B: 18.4 (#290), Llama 3.1-70B: 13.5 (#304)
| Benchmark | Llama 3.1-405B | Llama 3.1-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 9.7% | 3.6% |
| Omni-MATH | 24.9% | 21% |
| LMArena Math | 1281 | 1252 |
| MATH Level 5 | 49.8% | 36.7% |
Knowledge Llama 3.1-405B leads
Llama 3.1-405B: 30.4 (#227), Llama 3.1-70B: 24.2 (#269)
| Benchmark | Llama 3.1-405B | Llama 3.1-70B |
|---|---|---|
| GPQA Diamond | 50.9% | 44.2% |
| MMLU-Pro | 72.3% | 65.3% |
| GPQA (HELM) | 52.2% | 42.6% |
| LMArena Expert | 1243 | 1209 |
| MMLU | 84.5% | 80.1% |
| Confabulations | 17.6% | — |
| ARC (AI2) Challenge | 95.3% | — |
| TriviaQA | 82.7% | — |
Multilingual Llama 3.1-405B leads
Llama 3.1-405B: 40.7 (#214), Llama 3.1-70B: 38.8 (#225)
| Benchmark | Llama 3.1-405B | Llama 3.1-70B |
|---|---|---|
| LMArena Non-English | 1248 | 1219 |
| LMArena Chinese | 1242 | 1215 |
| LMArena French | 1279 | 1261 |
| LMArena German | 1252 | 1222 |
| LMArena Japanese | 1208 | 1132 |
| LMArena Korean | 1184 | 1140 |
| LMArena Russian | 1265 | 1234 |
| LMArena Spanish | 1260 | 1253 |
Instruction Following Too close to call
Llama 3.1-405B: 65.9 (#214), Llama 3.1-70B: 65.3 (#223)
| Benchmark | Llama 3.1-405B | Llama 3.1-70B |
|---|---|---|
| IFEval | 81.1% | 82.1% |
| LMArena Instruction Following | 1259 | 1231 |
Long Context Too close to call
Llama 3.1-405B: 38.4 (#197), Llama 3.1-70B: 37.6 (#214)
| Benchmark | Llama 3.1-405B | Llama 3.1-70B |
|---|---|---|
| LMArena Longer Query | 1266 | 1241 |
Writing & Preference Llama 3.1-405B leads
Llama 3.1-405B: 38.9 (#251), Llama 3.1-70B: 35.4 (#267)
| Benchmark | Llama 3.1-405B | Llama 3.1-70B |
|---|---|---|
| LMArena Text | 1284 | 1261 |
| LMArena Creative Writing | 1262 | 1232 |
| EQ-Bench Creative Writing | 870 | 784 |
| WildBench | 78.3% | 75.8% |
| LMArena Multi-Turn | 1297 | 1256 |
Frequently asked questions
Is Llama 3.1-405B better than Llama 3.1-70B?
Llama 3.1-405B is the stronger model overall, scoring 30.7 to 29.6 on the Noometry Index.
Is Llama 3.1-405B or Llama 3.1-70B better for coding?
Llama 3.1-405B scores higher on coding benchmarks: 33.1 versus 30.3 in the Noometry coding category.
How many benchmarks do Llama 3.1-405B and Llama 3.1-70B share?
31 benchmarks have published results for both models. Llama 3.1-405B has 42 scored results on Noometry and Llama 3.1-70B has 35.