Model comparison
Llama 2-7B vs Llama 3-8B
Llama 2-7B is the stronger model overall, scoring 29.1 to 25.5 on the Noometry Index.
Last verified . 22 shared benchmarks.
MetaSummary
- They share 22 benchmarks with published results for both. Llama 2-7B scores higher in 3 categories and Llama 3-8B in 5 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Llama 2-7B leads 30.7 to 8.8.
Side by side
| Llama 2-7B | Llama 3-8B | |
|---|---|---|
| Provider | Meta | Meta |
| Noometry Index | 29.1 | 25.5 |
| Released | 2023-07-18 | 2024-04-18 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 29 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 3-8B leads
Llama 2-7B: 29.2 (#307), Llama 3-8B: 31.0 (#289)
| Benchmark | Llama 2-7B | Llama 3-8B |
|---|---|---|
| LMArena Coding | 1002 | 1152 |
| BigCodeBench Instruct | — | 31.9% |
| BigCodeBench Complete | — | 36.9% |
| HumanEval+ | — | 56.7% |
| MBPP+ | — | 54.8% |
Reasoning Llama 2-7B leads
Llama 2-7B: 15.7 (#312), Llama 3-8B: 14.3 (#326)
| Benchmark | Llama 2-7B | Llama 3-8B |
|---|---|---|
| Chess Puzzles | 0% | 0% |
| LMArena Hard Prompts | 1009 | 1133 |
| Epoch Capabilities Index | 99.06 | 116.45 |
| WinoGrande | 69.2% | 75.7% |
| DTBench | — | 43.9% |
| Adversarial NLI | — | 57.3% |
| BIG-Bench Hard | 39.2% | — |
| ForecastBench | — | 58.6 |
| HellaSwag | 77.2% | — |
| LAMBADA | 73.3% | — |
| PIQA | 78.8% | — |
Math Llama 2-7B leads
Llama 2-7B: 30.7 (#233), Llama 3-8B: 8.8 (#323)
| Benchmark | Llama 2-7B | Llama 3-8B |
|---|---|---|
| LMArena Math | 1042 | 1151 |
| OTIS Mock AIME 2024-2025 | — | 1.9% |
| MATH Level 5 | — | 6.1% |
| GSM8K | 16.7% | — |
Knowledge Llama 2-7B leads
Llama 2-7B: 28.2 (#248), Llama 3-8B: 7.8 (#308)
| Benchmark | Llama 2-7B | Llama 3-8B |
|---|---|---|
| LMArena Expert | 1036 | 1113 |
| ARC (AI2) Challenge | 45.9% | 82.8% |
| MMLU | 45.8% | 68.8% |
| OpenBookQA | 58.6% | 82.6% |
| TriviaQA | 73.7% | 67.7% |
| GPQA Diamond | — | 26.1% |
| BoolQ | 77.9% | — |
Multimodal Not comparable
Llama 2-7B: —, Llama 3-8B: —
| Benchmark | Llama 2-7B | Llama 3-8B |
|---|---|---|
| ScienceQA | 43.1% | — |
Multilingual Llama 3-8B leads
Llama 2-7B: 23.8 (#293), Llama 3-8B: 30.8 (#261)
| Benchmark | Llama 2-7B | Llama 3-8B |
|---|---|---|
| LMArena Non-English | 973 | 1098 |
| LMArena Chinese | 973 | 1076 |
| LMArena French | 970 | 1159 |
| LMArena German | 978 | 1104 |
| LMArena Russian | 995 | 1109 |
| LMArena Spanish | 1007 | 1173 |
| LMArena Japanese | — | 967 |
| LMArena Korean | — | 1004 |
Instruction Following Llama 3-8B leads
Llama 2-7B: 50.8 (#298), Llama 3-8B: 58.4 (#260)
| Benchmark | Llama 2-7B | Llama 3-8B |
|---|---|---|
| LMArena Instruction Following | 1006 | 1127 |
Long Context Llama 3-8B leads
Llama 2-7B: 30.4 (#287), Llama 3-8B: 34.2 (#251)
| Benchmark | Llama 2-7B | Llama 3-8B |
|---|---|---|
| LMArena Longer Query | 999 | 1128 |
Writing & Preference Llama 3-8B leads
Llama 2-7B: 28.0 (#298), Llama 3-8B: 37.5 (#256)
| Benchmark | Llama 2-7B | Llama 3-8B |
|---|---|---|
| LMArena Text | 1053 | 1166 |
| LMArena Creative Writing | 1033 | 1150 |
| LMArena Multi-Turn | 1029 | 1152 |
Frequently asked questions
Is Llama 2-7B better than Llama 3-8B?
Llama 2-7B is the stronger model overall, scoring 29.1 to 25.5 on the Noometry Index.
Is Llama 2-7B or Llama 3-8B better for coding?
Llama 3-8B scores higher on coding benchmarks: 31.0 versus 29.2 in the Noometry coding category.
How many benchmarks do Llama 2-7B and Llama 3-8B share?
22 benchmarks have published results for both models. Llama 2-7B has 29 scored results on Noometry and Llama 3-8B has 34.