Model comparison
Gemma 2 9B vs Llama-3.3-70B-Instruct
Llama-3.3-70B-Instruct is the stronger model overall, scoring 30.3 to 25.7 on the Noometry Index.
Last verified . 31 shared benchmarks.
Google
MetaSummary
- They share 31 benchmarks with published results for both. Gemma 2 9B scores higher in 2 categories and Llama-3.3-70B-Instruct in 6 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Llama-3.3-70B-Instruct leads 30.5 to 9.7.
- The biggest single-benchmark swing is LiveBench Reasoning: 15.2% for Gemma 2 9B and 50.8% for Llama-3.3-70B-Instruct.
Side by side
| Gemma 2 9B | Llama-3.3-70B-Instruct | |
|---|---|---|
| Provider | Meta | |
| Noometry Index | 25.7 | 30.3 |
| Released | 2024-06-24 | 2024-12-06 |
| Weights | Open | Open |
| Context window | — | 128K |
| Max output | — | 4K |
| Input $ / M tokens | — | $0.10 |
| Output $ / M tokens | — | $0.32 |
| Results tracked | 35 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama-3.3-70B-Instruct leads
Gemma 2 9B: 29.0 (#299), Llama-3.3-70B-Instruct: 30.6 (#284)
| Benchmark | Gemma 2 9B | Llama-3.3-70B-Instruct |
|---|---|---|
| BigCodeBench Instruct | 34.7% | 46.9% |
| LiveBench Coding | 22.5% | 36.6% |
| LMArena Coding | 1174 | 1268 |
| BigCodeBench Complete | 40.6% | 57.5% |
| SciCode | — | 26% |
| WeirdML | — | 14.4% |
Agentic & Tool Use Not comparable
Gemma 2 9B: —, Llama-3.3-70B-Instruct: 25.6 (#100)
| Benchmark | Gemma 2 9B | Llama-3.3-70B-Instruct |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 31.9% |
| BALROG | — | 23% |
Reasoning Gemma 2 9B leads
Gemma 2 9B: 15.3 (#306), Llama-3.3-70B-Instruct: 13.7 (#325)
| Benchmark | Gemma 2 9B | Llama-3.3-70B-Instruct |
|---|---|---|
| LiveBench Reasoning | 15.2% | 50.8% |
| LMArena Hard Prompts | 1172 | 1257 |
| LiveBench Data Analysis | 36.4% | 49.5% |
| Epoch Capabilities Index | 119.78 | 127.32 |
| LiveBench | 28.7% | 50.2% |
| SimpleBench | — | 19.9% |
| CritPt | — | 0% |
| DTBench | — | 59.5% |
| LMCA | — | 17.5% |
| ForecastBench | — | 58.6 |
| PIQA | 83.7% | — |
Math Llama-3.3-70B-Instruct leads
Gemma 2 9B: 9.3 (#312), Llama-3.3-70B-Instruct: 14.3 (#292)
| Benchmark | Gemma 2 9B | Llama-3.3-70B-Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 0.6% | 5.1% |
| LiveBench Math | 19.8% | 42.2% |
| LMArena Math | 1183 | 1268 |
| MATH Level 5 | 21% | 41.6% |
| GSM8K | 84.9% | — |
Knowledge Llama-3.3-70B-Instruct leads
Gemma 2 9B: 9.7 (#300), Llama-3.3-70B-Instruct: 30.5 (#220)
| Benchmark | Gemma 2 9B | Llama-3.3-70B-Instruct |
|---|---|---|
| GPQA Diamond | 27.5% | 47.4% |
| LMArena Expert | 1147 | 1225 |
| MMLU | 72.1% | 86.3% |
| Confabulations | — | 22.8% |
| Vectara Hallucination Rate | — | 4.1% |
| BoolQ | 85.7% | — |
Multilingual Llama-3.3-70B-Instruct leads
Gemma 2 9B: 36.8 (#233), Llama-3.3-70B-Instruct: 40.1 (#215)
| Benchmark | Gemma 2 9B | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Non-English | 1188 | 1236 |
| LMArena Chinese | 1185 | 1217 |
| LMArena French | 1189 | 1280 |
| LMArena German | 1186 | 1251 |
| LMArena Japanese | 1144 | 1149 |
| LMArena Korean | 1137 | 1142 |
| LMArena Russian | 1200 | 1252 |
| LMArena Spanish | 1200 | 1269 |
Instruction Following Llama-3.3-70B-Instruct leads
Gemma 2 9B: 57.8 (#264), Llama-3.3-70B-Instruct: 71.3 (#152)
| Benchmark | Gemma 2 9B | Llama-3.3-70B-Instruct |
|---|---|---|
| LiveBench Instruction Following | 52.6% | 82.7% |
| LMArena Instruction Following | 1178 | 1242 |
Long Context Gemma 2 9B leads
Gemma 2 9B: 36.3 (#228), Llama-3.3-70B-Instruct: 26.4 (#290)
| Benchmark | Gemma 2 9B | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Longer Query | 1197 | 1256 |
| Fiction.LiveBench | — | 33.3% |
Writing & Preference Llama-3.3-70B-Instruct leads
Gemma 2 9B: 32.3 (#276), Llama-3.3-70B-Instruct: 47.8 (#203)
| Benchmark | Gemma 2 9B | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Text | 1207 | 1274 |
| LMArena Creative Writing | 1206 | 1250 |
| LMArena Multi-Turn | 1193 | 1279 |
| LiveBench Language | 25.5% | 39.2% |
| EQ-Bench Creative Writing | 841 | — |
Frequently asked questions
Is Gemma 2 9B better than Llama-3.3-70B-Instruct?
Llama-3.3-70B-Instruct is the stronger model overall, scoring 30.3 to 25.7 on the Noometry Index.
Is Gemma 2 9B or Llama-3.3-70B-Instruct better for coding?
Llama-3.3-70B-Instruct scores higher on coding benchmarks: 30.6 versus 29.0 in the Noometry coding category.
How many benchmarks do Gemma 2 9B and Llama-3.3-70B-Instruct share?
31 benchmarks have published results for both models. Gemma 2 9B has 35 scored results on Noometry and Llama-3.3-70B-Instruct has 43.