Model comparison
Llama 3.1-405B vs Llama-3.3-70B-Instruct
Llama 3.1-405B and Llama-3.3-70B-Instruct score almost the same on the Noometry Index (30.7 vs 30.6), 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. Llama 3.1-405B scores higher in 5 categories and Llama-3.3-70B-Instruct in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in long context, where Llama 3.1-405B leads 38.4 to 26.4.
- The biggest single-benchmark swing is MATH Level 5: 49.8% for Llama 3.1-405B and 41.6% for Llama-3.3-70B-Instruct.
Side by side
| Llama 3.1-405B | Llama-3.3-70B-Instruct | |
|---|---|---|
| Provider | Meta | Meta |
| Noometry Index | 30.7 | 30.6 |
| Released | 2024-07-23 | 2024-12-06 |
| Weights | Open | Open |
| Context window | — | 128K |
| Max output | — | 4K |
| Input $ / M tokens | — | $0.10 |
| Output $ / M tokens | — | $0.32 |
| Results tracked | 42 | 43 |
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-3.3-70B-Instruct: 31.0 (#290)
| Benchmark | Llama 3.1-405B | Llama-3.3-70B-Instruct |
|---|---|---|
| WeirdML | 21.4% | 14.4% |
| LMArena Coding | 1291 | 1268 |
| SciCode | — | 26% |
| BigCodeBench Instruct | — | 46.9% |
| LiveBench Coding | — | 36.6% |
| BigCodeBench Complete | — | 57.5% |
Agentic & Tool Use Llama-3.3-70B-Instruct leads
Llama 3.1-405B: 21.0 (#140), Llama-3.3-70B-Instruct: 25.8 (#105)
| Benchmark | Llama 3.1-405B | Llama-3.3-70B-Instruct |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 31.9% |
| TheAgentCompany | 7.4% | — |
| Cybench | 7.5% | — |
| BALROG | — | 23% |
Reasoning Llama 3.1-405B leads
Llama 3.1-405B: 16.8 (#300), Llama-3.3-70B-Instruct: 14.1 (#327)
| Benchmark | Llama 3.1-405B | Llama-3.3-70B-Instruct |
|---|---|---|
| SimpleBench | 23% | 19.9% |
| LMArena Hard Prompts | 1269 | 1257 |
| DTBench | 61.4% | 59.5% |
| Epoch Capabilities Index | 128.75 | 127.33 |
| ForecastBench | 59.9 | 58.6 |
| Kagi LLM Benchmark | 45% | — |
| CritPt | — | 0% |
| LiveBench Reasoning | — | 50.8% |
| LiveBench Data Analysis | — | 49.5% |
| LMCA | — | 17.5% |
| BIG-Bench Hard | 82.9% | — |
| HellaSwag | 89.2% | — |
| LiveBench | — | 50.2% |
| PIQA | 85.9% | — |
| WinoGrande | 89.2% | — |
Math Llama 3.1-405B leads
Llama 3.1-405B: 18.4 (#290), Llama-3.3-70B-Instruct: 15.3 (#298)
| Benchmark | Llama 3.1-405B | Llama-3.3-70B-Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 9.7% | 5.1% |
| LMArena Math | 1281 | 1267 |
| MATH Level 5 | 49.8% | 41.6% |
| Omni-MATH | 24.9% | — |
| LiveBench Math | — | 42.2% |
Knowledge Too close to call
Llama 3.1-405B: 30.4 (#227), Llama-3.3-70B-Instruct: 30.6 (#226)
| Benchmark | Llama 3.1-405B | Llama-3.3-70B-Instruct |
|---|---|---|
| GPQA Diamond | 50.9% | 47.4% |
| Confabulations | 17.6% | 22.8% |
| LMArena Expert | 1243 | 1225 |
| MMLU | 84.5% | 86.3% |
| MMLU-Pro | 72.3% | — |
| Vectara Hallucination Rate | — | 4.1% |
| GPQA (HELM) | 52.2% | — |
| ARC (AI2) Challenge | 95.3% | — |
| TriviaQA | 82.7% | — |
Multilingual Too close to call
Llama 3.1-405B: 40.7 (#214), Llama-3.3-70B-Instruct: 39.9 (#220)
| Benchmark | Llama 3.1-405B | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Non-English | 1248 | 1236 |
| LMArena Chinese | 1242 | 1217 |
| LMArena French | 1279 | 1281 |
| LMArena German | 1252 | 1251 |
| LMArena Japanese | 1208 | 1150 |
| LMArena Korean | 1184 | 1143 |
| LMArena Russian | 1265 | 1252 |
| LMArena Spanish | 1260 | 1270 |
Instruction Following Llama-3.3-70B-Instruct leads
Llama 3.1-405B: 65.9 (#214), Llama-3.3-70B-Instruct: 71.1 (#157)
| Benchmark | Llama 3.1-405B | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Instruction Following | 1259 | 1242 |
| LiveBench Instruction Following | — | 82.7% |
| IFEval | 81.1% | — |
Long Context Llama 3.1-405B leads
Llama 3.1-405B: 38.4 (#197), Llama-3.3-70B-Instruct: 26.4 (#295)
| Benchmark | Llama 3.1-405B | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Longer Query | 1266 | 1256 |
| Fiction.LiveBench | — | 33.3% |
Writing & Preference Llama-3.3-70B-Instruct leads
Llama 3.1-405B: 38.9 (#251), Llama-3.3-70B-Instruct: 47.6 (#207)
| Benchmark | Llama 3.1-405B | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Text | 1284 | 1274 |
| LMArena Creative Writing | 1262 | 1250 |
| LMArena Multi-Turn | 1297 | 1280 |
| EQ-Bench Creative Writing | 870 | — |
| WildBench | 78.3% | — |
| LiveBench Language | — | 39.2% |
Frequently asked questions
Is Llama 3.1-405B better than Llama-3.3-70B-Instruct?
Llama 3.1-405B and Llama-3.3-70B-Instruct score almost the same on the Noometry Index (30.7 vs 30.6), so choose on price, context window or the category you care about most.
Is Llama 3.1-405B or Llama-3.3-70B-Instruct better for coding?
Llama 3.1-405B scores higher on coding benchmarks: 33.1 versus 31.0 in the Noometry coding category.
How many benchmarks do Llama 3.1-405B and Llama-3.3-70B-Instruct share?
27 benchmarks have published results for both models. Llama 3.1-405B has 42 scored results on Noometry and Llama-3.3-70B-Instruct has 43.