Model comparison
Gemma 2 9B vs Llama 2-70B
Gemma 2 9B is the stronger model overall, scoring 25.7 to 24.1 on the Noometry Index.
Last verified . 25 shared benchmarks.
Google
MetaSummary
- They share 25 benchmarks with published results for both. Gemma 2 9B scores higher in 6 categories and Llama 2-70B in 2 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in multilingual, where Gemma 2 9B leads 36.8 to 27.9.
- The biggest single-benchmark swing is MATH Level 5: 21% for Gemma 2 9B and 3.3% for Llama 2-70B.
Side by side
| Gemma 2 9B | Llama 2-70B | |
|---|---|---|
| Provider | Meta | |
| Noometry Index | 25.7 | 24.1 |
| Released | 2024-06-24 | 2023-07-18 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 35 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 2-70B leads
Gemma 2 9B: 29.0 (#299), Llama 2-70B: 30.8 (#281)
| Benchmark | Gemma 2 9B | Llama 2-70B |
|---|---|---|
| LMArena Coding | 1174 | 1079 |
| BigCodeBench Instruct | 34.7% | — |
| LiveBench Coding | 22.5% | — |
| BigCodeBench Complete | 40.6% | — |
Reasoning Gemma 2 9B leads
Gemma 2 9B: 15.3 (#306), Llama 2-70B: 13.7 (#324)
| Benchmark | Gemma 2 9B | Llama 2-70B |
|---|---|---|
| LMArena Hard Prompts | 1172 | 1073 |
| Epoch Capabilities Index | 119.78 | 113.59 |
| PIQA | 83.7% | 82.8% |
| LiveBench Reasoning | 15.2% | — |
| DTBench | — | 41.6% |
| LiveBench Data Analysis | 36.4% | — |
| BIG-Bench Hard | — | 64.9% |
| CommonsenseQA 2.0 | — | 50% |
| ForecastBench | — | 51.4 |
| HellaSwag | — | 85.3% |
| LAMBADA | — | 78.9% |
| LiveBench | 28.7% | — |
| WinoGrande | — | 80.2% |
Math Gemma 2 9B leads
Gemma 2 9B: 9.3 (#312), Llama 2-70B: 7.5 (#320)
| Benchmark | Gemma 2 9B | Llama 2-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 0.6% | 0% |
| LMArena Math | 1183 | 1091 |
| MATH Level 5 | 21% | 3.3% |
| GSM8K | 84.9% | 69.6% |
| LiveBench Math | 19.8% | — |
Knowledge Gemma 2 9B leads
Gemma 2 9B: 9.7 (#300), Llama 2-70B: 7.4 (#305)
| Benchmark | Gemma 2 9B | Llama 2-70B |
|---|---|---|
| GPQA Diamond | 27.5% | 26.3% |
| LMArena Expert | 1147 | 1039 |
| BoolQ | 85.7% | 88.6% |
| MMLU | 72.1% | 69.9% |
| ARC (AI2) Challenge | — | 78.3% |
| OpenBookQA | — | 60.2% |
| TriviaQA | — | 87.6% |
Multilingual Gemma 2 9B leads
Gemma 2 9B: 36.8 (#233), Llama 2-70B: 27.9 (#269)
| Benchmark | Gemma 2 9B | Llama 2-70B |
|---|---|---|
| LMArena Non-English | 1188 | 1045 |
| LMArena Chinese | 1185 | 994 |
| LMArena French | 1189 | 1089 |
| LMArena German | 1186 | 1041 |
| LMArena Japanese | 1144 | 927 |
| LMArena Korean | 1137 | 964 |
| LMArena Russian | 1200 | 1083 |
| LMArena Spanish | 1200 | 1142 |
Instruction Following Gemma 2 9B leads
Gemma 2 9B: 57.8 (#264), Llama 2-70B: 55.1 (#273)
| Benchmark | Gemma 2 9B | Llama 2-70B |
|---|---|---|
| LMArena Instruction Following | 1178 | 1071 |
| LiveBench Instruction Following | 52.6% | — |
Long Context Gemma 2 9B leads
Gemma 2 9B: 36.3 (#228), Llama 2-70B: 32.2 (#265)
| Benchmark | Gemma 2 9B | Llama 2-70B |
|---|---|---|
| LMArena Longer Query | 1197 | 1062 |
Writing & Preference Too close to call
Gemma 2 9B: 32.3 (#276), Llama 2-70B: 32.5 (#274)
| Benchmark | Gemma 2 9B | Llama 2-70B |
|---|---|---|
| LMArena Text | 1207 | 1115 |
| LMArena Creative Writing | 1206 | 1075 |
| LMArena Multi-Turn | 1193 | 1088 |
| EQ-Bench Creative Writing | 841 | — |
| LiveBench Language | 25.5% | — |
Frequently asked questions
Is Gemma 2 9B better than Llama 2-70B?
Gemma 2 9B is the stronger model overall, scoring 25.7 to 24.1 on the Noometry Index.
Is Gemma 2 9B or Llama 2-70B better for coding?
Llama 2-70B scores higher on coding benchmarks: 30.8 versus 29.0 in the Noometry coding category.
How many benchmarks do Gemma 2 9B and Llama 2-70B share?
25 benchmarks have published results for both models. Gemma 2 9B has 35 scored results on Noometry and Llama 2-70B has 35.