Model comparison
Gemma 7B vs Llama 2-70B
Gemma 7B is the stronger model overall, scoring 30.0 to 24.4 on the Noometry Index.
Last verified . 24 shared benchmarks.
Google
MetaSummary
- They share 24 benchmarks with published results for both. Gemma 7B scores higher in 3 categories and Llama 2-70B in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Gemma 7B leads 31.2 to 8.1.
Side by side
| Gemma 7B | Llama 2-70B | |
|---|---|---|
| Provider | Meta | |
| Noometry Index | 30.0 | 24.4 |
| Released | 2024-02-21 | 2023-07-18 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 27 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Gemma 7B: 30.5 (#294), Llama 2-70B: 31.4 (#286)
| Benchmark | Gemma 7B | Llama 2-70B |
|---|---|---|
| LMArena Coding | 1048 | 1079 |
| HumanEval+ | 28.7% | — |
| MBPP+ | 43.4% | — |
Reasoning Gemma 7B leads
Gemma 7B: 19.9 (#249), Llama 2-70B: 14.4 (#325)
| Benchmark | Gemma 7B | Llama 2-70B |
|---|---|---|
| LMArena Hard Prompts | 1042 | 1073 |
| BIG-Bench Hard | 55.1% | 64.9% |
| Epoch Capabilities Index | 111.99 | 113.79 |
| HellaSwag | 82.2% | 85.3% |
| PIQA | 81.2% | 82.8% |
| WinoGrande | 79% | 80.2% |
| DTBench | — | 41.6% |
| Adversarial NLI | 48.7% | — |
| CommonsenseQA 2.0 | — | 50% |
| ForecastBench | — | 51.4 |
| LAMBADA | — | 78.9% |
Math Gemma 7B leads
Gemma 7B: 31.2 (#228), Llama 2-70B: 8.1 (#326)
| Benchmark | Gemma 7B | Llama 2-70B |
|---|---|---|
| LMArena Math | 1066 | 1091 |
| GSM8K | 46.4% | 69.6% |
| OTIS Mock AIME 2024-2025 | — | 0% |
| MATH Level 5 | — | 3.3% |
Knowledge Gemma 7B leads
Gemma 7B: 27.3 (#252), Llama 2-70B: 7.4 (#310)
| Benchmark | Gemma 7B | Llama 2-70B |
|---|---|---|
| LMArena Expert | 1001 | 1039 |
| ARC (AI2) Challenge | 78.3% | 78.3% |
| BoolQ | 83.2% | 88.6% |
| MMLU | 66.1% | 69.9% |
| OpenBookQA | 78.6% | 60.2% |
| TriviaQA | 72.3% | 87.6% |
| GPQA Diamond | — | 26.3% |
Multilingual Llama 2-70B leads
Gemma 7B: 25.1 (#287), Llama 2-70B: 27.7 (#274)
| Benchmark | Gemma 7B | Llama 2-70B |
|---|---|---|
| LMArena Non-English | 999 | 1045 |
| LMArena Chinese | 1035 | 995 |
| LMArena French | 1025 | 1090 |
| LMArena Russian | 993 | 1083 |
| LMArena German | — | 1041 |
| LMArena Japanese | — | 927 |
| LMArena Korean | — | 964 |
| LMArena Spanish | — | 1143 |
Instruction Following Llama 2-70B leads
Gemma 7B: 51.5 (#295), Llama 2-70B: 54.9 (#278)
| Benchmark | Gemma 7B | Llama 2-70B |
|---|---|---|
| LMArena Instruction Following | 1017 | 1071 |
Long Context Llama 2-70B leads
Gemma 7B: 31.1 (#282), Llama 2-70B: 32.3 (#270)
| Benchmark | Gemma 7B | Llama 2-70B |
|---|---|---|
| LMArena Longer Query | 1022 | 1062 |
Writing & Preference Llama 2-70B leads
Gemma 7B: 27.1 (#302), Llama 2-70B: 32.3 (#279)
| Benchmark | Gemma 7B | Llama 2-70B |
|---|---|---|
| LMArena Text | 1056 | 1115 |
| LMArena Creative Writing | 1024 | 1075 |
| LMArena Multi-Turn | 963 | 1088 |
Frequently asked questions
Is Gemma 7B better than Llama 2-70B?
Gemma 7B is the stronger model overall, scoring 30.0 to 24.4 on the Noometry Index.
Is Gemma 7B or Llama 2-70B better for coding?
They score almost the same on coding (30.5 vs 31.4); test both on your own repository before choosing.
How many benchmarks do Gemma 7B and Llama 2-70B share?
24 benchmarks have published results for both models. Gemma 7B has 27 scored results on Noometry and Llama 2-70B has 35.