Model comparison
Llama 2-70B vs Llama 4 Scout
Llama 4 Scout is the stronger model overall, scoring 27.4 to 24.1 on the Noometry Index.
Last verified . 23 shared benchmarks.
MetaSummary
- They share 23 benchmarks with published results for both. Llama 2-70B scores higher in 3 categories and Llama 4 Scout in 5 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Llama 4 Scout leads 31.8 to 7.4.
- The biggest single-benchmark swing is MATH Level 5: 3.3% for Llama 2-70B and 62.3% for Llama 4 Scout.
Side by side
| Llama 2-70B | Llama 4 Scout | |
|---|---|---|
| Provider | Meta | Meta |
| Noometry Index | 24.1 | 27.4 |
| Released | 2023-07-18 | 2025-04-05 |
| Weights | Open | Open |
| Context window | — | 128K |
| Max output | — | 4K |
| Input $ / M tokens | — | $0.10 |
| Output $ / M tokens | — | $0.30 |
| Results tracked | 35 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 2-70B leads
Llama 2-70B: 30.8 (#281), Llama 4 Scout: 19.9 (#334)
| Benchmark | Llama 2-70B | Llama 4 Scout |
|---|---|---|
| LMArena Coding | 1079 | 1286 |
| SWE-bench Verified (bash only) | — | 9.1% |
| SciCode | — | 17% |
| BigCodeBench Complete | — | 43.1% |
Agentic & Tool Use Not comparable
Llama 2-70B: —, Llama 4 Scout: 24.4 (#109)
| Benchmark | Llama 2-70B | Llama 4 Scout |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 28.1% |
Reasoning Llama 2-70B leads
Llama 2-70B: 13.7 (#324), Llama 4 Scout: 9.2 (#339)
| Benchmark | Llama 2-70B | Llama 4 Scout |
|---|---|---|
| LMArena Hard Prompts | 1073 | 1266 |
| DTBench | 41.6% | 57.9% |
| Epoch Capabilities Index | 113.59 | 129.64 |
| ForecastBench | 51.4 | 57.5 |
| ARC-AGI-2 | — | 0% |
| Kagi LLM Benchmark | — | 36.9% |
| ARC-AGI-1 | — | 0.5% |
| CritPt | — | 0% |
| LMCA | — | 12% |
| BIG-Bench Hard | 64.9% | — |
| CommonsenseQA 2.0 | 50% | — |
| HellaSwag | 85.3% | — |
| LAMBADA | 78.9% | — |
| PIQA | 82.8% | — |
| WinoGrande | 80.2% | — |
Math Llama 4 Scout leads
Llama 2-70B: 7.5 (#320), Llama 4 Scout: 18.4 (#280)
| Benchmark | Llama 2-70B | Llama 4 Scout |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 0% | 7.8% |
| LMArena Math | 1091 | 1286 |
| MATH Level 5 | 3.3% | 62.3% |
| Omni-MATH | — | 37.3% |
| FrontierMath (Feb 2025 set) | — | 0% |
| GSM8K | 69.6% | — |
Knowledge Llama 4 Scout leads
Llama 2-70B: 7.4 (#305), Llama 4 Scout: 31.8 (#208)
| Benchmark | Llama 2-70B | Llama 4 Scout |
|---|---|---|
| GPQA Diamond | 26.3% | 51.8% |
| LMArena Expert | 1039 | 1236 |
| MMLU-Pro | — | 74.2% |
| Vectara Hallucination Rate | — | 7.7% |
| GPQA (HELM) | — | 50.7% |
| ARC (AI2) Challenge | 78.3% | — |
| BoolQ | 88.6% | — |
| MMLU | 69.9% | — |
| OpenBookQA | 60.2% | — |
| TriviaQA | 87.6% | — |
Multimodal Not comparable
Llama 2-70B: —, Llama 4 Scout: 31.6 (#98)
| Benchmark | Llama 2-70B | Llama 4 Scout |
|---|---|---|
| LMArena Vision | — | 1118 |
| SpatialViz-Bench | — | 34.2% |
Multilingual Llama 4 Scout leads
Llama 2-70B: 27.9 (#269), Llama 4 Scout: 41.3 (#207)
| Benchmark | Llama 2-70B | Llama 4 Scout |
|---|---|---|
| LMArena Non-English | 1045 | 1253 |
| LMArena Chinese | 994 | 1256 |
| LMArena French | 1089 | 1283 |
| LMArena German | 1041 | 1274 |
| LMArena Japanese | 927 | 1209 |
| LMArena Korean | 964 | 1207 |
| LMArena Russian | 1083 | 1263 |
| LMArena Spanish | 1142 | 1277 |
Instruction Following Llama 4 Scout leads
Llama 2-70B: 55.1 (#273), Llama 4 Scout: 66.0 (#212)
| Benchmark | Llama 2-70B | Llama 4 Scout |
|---|---|---|
| LMArena Instruction Following | 1071 | 1248 |
| IFEval | — | 81.8% |
Long Context Llama 2-70B leads
Llama 2-70B: 32.2 (#265), Llama 4 Scout: 27.5 (#289)
| Benchmark | Llama 2-70B | Llama 4 Scout |
|---|---|---|
| LMArena Longer Query | 1062 | 1265 |
| Fiction.LiveBench | — | 36% |
Writing & Preference Llama 4 Scout leads
Llama 2-70B: 32.5 (#274), Llama 4 Scout: 37.1 (#256)
| Benchmark | Llama 2-70B | Llama 4 Scout |
|---|---|---|
| LMArena Text | 1115 | 1279 |
| LMArena Creative Writing | 1075 | 1249 |
| LMArena Multi-Turn | 1088 | 1281 |
| EQ-Bench Creative Writing | — | 783 |
| WildBench | — | 78% |
Frequently asked questions
Is Llama 2-70B better than Llama 4 Scout?
Llama 4 Scout is the stronger model overall, scoring 27.4 to 24.1 on the Noometry Index.
Is Llama 2-70B or Llama 4 Scout better for coding?
Llama 2-70B scores higher on coding benchmarks: 30.8 versus 19.9 in the Noometry coding category.
How many benchmarks do Llama 2-70B and Llama 4 Scout share?
23 benchmarks have published results for both models. Llama 2-70B has 35 scored results on Noometry and Llama 4 Scout has 43.