Model comparison
Gemma 2B vs Llama 2-70B
Gemma 2B is the stronger model overall, scoring 29.6 to 24.4 on the Noometry Index.
Last verified . 21 shared benchmarks.
Google
MetaSummary
- They share 21 benchmarks with published results for both. Gemma 2B scores higher in 2 categories and Llama 2-70B in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Gemma 2B leads 30.0 to 8.1.
Side by side
| Gemma 2B | Llama 2-70B | |
|---|---|---|
| Provider | Meta | |
| Noometry Index | 29.6 | 24.4 |
| Released | 2024-02-21 | 2023-07-18 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 23 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 2-70B leads
Gemma 2B: 29.4 (#305), Llama 2-70B: 31.4 (#286)
| Benchmark | Gemma 2B | Llama 2-70B |
|---|---|---|
| LMArena Coding | 1010 | 1079 |
| HumanEval+ | 20.7% | — |
| MBPP+ | 34.1% | — |
Reasoning Gemma 2B leads
Gemma 2B: 18.8 (#275), Llama 2-70B: 14.4 (#325)
| Benchmark | Gemma 2B | Llama 2-70B |
|---|---|---|
| LMArena Hard Prompts | 989 | 1073 |
| BIG-Bench Hard | 35.2% | 64.9% |
| Epoch Capabilities Index | 94.2 | 113.79 |
| HellaSwag | 71.4% | 85.3% |
| PIQA | 77.3% | 82.8% |
| WinoGrande | 65.4% | 80.2% |
| DTBench | — | 41.6% |
| CommonsenseQA 2.0 | — | 50% |
| ForecastBench | — | 51.4 |
| LAMBADA | — | 78.9% |
Math Gemma 2B leads
Gemma 2B: 30.0 (#239), Llama 2-70B: 8.1 (#326)
| Benchmark | Gemma 2B | Llama 2-70B |
|---|---|---|
| LMArena Math | 1009 | 1091 |
| GSM8K | 17.7% | 69.6% |
| OTIS Mock AIME 2024-2025 | — | 0% |
| MATH Level 5 | — | 3.3% |
Knowledge Not comparable
Gemma 2B: —, Llama 2-70B: 7.4 (#310)
| Benchmark | Gemma 2B | Llama 2-70B |
|---|---|---|
| ARC (AI2) Challenge | 42.1% | 78.3% |
| BoolQ | 69.4% | 88.6% |
| MMLU | 42.3% | 69.9% |
| TriviaQA | 53.2% | 87.6% |
| GPQA Diamond | — | 26.3% |
| LMArena Expert | — | 1039 |
| OpenBookQA | — | 60.2% |
Multilingual Llama 2-70B leads
Gemma 2B: 23.0 (#294), Llama 2-70B: 27.7 (#274)
| Benchmark | Gemma 2B | Llama 2-70B |
|---|---|---|
| LMArena Non-English | 958 | 1045 |
| LMArena Chinese | 986 | 995 |
| LMArena Russian | 937 | 1083 |
| LMArena French | — | 1090 |
| LMArena German | — | 1041 |
| LMArena Japanese | — | 927 |
| LMArena Korean | — | 964 |
| LMArena Spanish | — | 1143 |
Instruction Following Llama 2-70B leads
Gemma 2B: 48.5 (#302), Llama 2-70B: 54.9 (#278)
| Benchmark | Gemma 2B | Llama 2-70B |
|---|---|---|
| LMArena Instruction Following | 970 | 1071 |
Long Context Llama 2-70B leads
Gemma 2B: 29.9 (#291), Llama 2-70B: 32.3 (#270)
| Benchmark | Gemma 2B | Llama 2-70B |
|---|---|---|
| LMArena Longer Query | 981 | 1062 |
Writing & Preference Llama 2-70B leads
Gemma 2B: 24.0 (#308), Llama 2-70B: 32.3 (#279)
| Benchmark | Gemma 2B | Llama 2-70B |
|---|---|---|
| LMArena Text | 1002 | 1115 |
| LMArena Creative Writing | 987 | 1075 |
| LMArena Multi-Turn | 945 | 1088 |
Frequently asked questions
Is Gemma 2B better than Llama 2-70B?
Gemma 2B is the stronger model overall, scoring 29.6 to 24.4 on the Noometry Index.
Is Gemma 2B or Llama 2-70B better for coding?
Llama 2-70B scores higher on coding benchmarks: 31.4 versus 29.4 in the Noometry coding category.
How many benchmarks do Gemma 2B and Llama 2-70B share?
21 benchmarks have published results for both models. Gemma 2B has 23 scored results on Noometry and Llama 2-70B has 35.