Model comparison
Gemma 7B vs Llama 3.2 3B
Gemma 7B is the stronger model overall, scoring 30.0 to 28.9 on the Noometry Index.
Last verified . 12 shared benchmarks.
Google
MetaSummary
- They share 12 benchmarks with published results for both. Gemma 7B scores higher in 2 categories and Llama 3.2 3B in 6 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in instruction following, where Llama 3.2 3B leads 56.0 to 51.5.
Side by side
| Gemma 7B | Llama 3.2 3B | |
|---|---|---|
| Provider | Meta | |
| Noometry Index | 30.0 | 28.9 |
| Released | 2024-02-21 | 2024-09-24 |
| Weights | Open | Open |
| Context window | — | 131K |
| Max output | — | 118K |
| Input $ / M tokens | — | $0.05 |
| Output $ / M tokens | — | $0.33 |
| Results tracked | 27 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 7B leads
Gemma 7B: 30.5 (#294), Llama 3.2 3B: 27.6 (#319)
| Benchmark | Gemma 7B | Llama 3.2 3B |
|---|---|---|
| LMArena Coding | 1048 | 1098 |
| BigCodeBench Instruct | — | 23.4% |
| BigCodeBench Complete | — | 28.3% |
| HumanEval+ | 28.7% | — |
| MBPP+ | 43.4% | — |
Agentic & Tool Use Not comparable
Gemma 7B: —, Llama 3.2 3B: 20.1 (#143)
| Benchmark | Gemma 7B | Llama 3.2 3B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 21.9% |
| BALROG | — | 10.1% |
Reasoning Llama 3.2 3B leads
Gemma 7B: 19.9 (#249), Llama 3.2 3B: 21.0 (#228)
| Benchmark | Gemma 7B | Llama 3.2 3B |
|---|---|---|
| LMArena Hard Prompts | 1042 | 1095 |
| Adversarial NLI | 48.7% | — |
| BIG-Bench Hard | 55.1% | — |
| Epoch Capabilities Index | 111.99 | — |
| HellaSwag | 82.2% | — |
| PIQA | 81.2% | — |
| WinoGrande | 79% | — |
Math Llama 3.2 3B leads
Gemma 7B: 31.2 (#228), Llama 3.2 3B: 32.4 (#214)
| Benchmark | Gemma 7B | Llama 3.2 3B |
|---|---|---|
| LMArena Math | 1066 | 1126 |
| GSM8K | 46.4% | — |
Knowledge Llama 3.2 3B leads
Gemma 7B: 27.3 (#252), Llama 3.2 3B: 29.7 (#235)
| Benchmark | Gemma 7B | Llama 3.2 3B |
|---|---|---|
| LMArena Expert | 1001 | 1090 |
| ARC (AI2) Challenge | 78.3% | — |
| BoolQ | 83.2% | — |
| MMLU | 66.1% | — |
| OpenBookQA | 78.6% | — |
| TriviaQA | 72.3% | — |
Multilingual Llama 3.2 3B leads
Gemma 7B: 25.1 (#287), Llama 3.2 3B: 26.2 (#281)
| Benchmark | Gemma 7B | Llama 3.2 3B |
|---|---|---|
| LMArena Non-English | 999 | 1019 |
| LMArena Chinese | 1035 | 1017 |
| LMArena Russian | 993 | 949 |
| LMArena French | 1025 | — |
| LMArena German | — | 1056 |
Instruction Following Llama 3.2 3B leads
Gemma 7B: 51.5 (#295), Llama 3.2 3B: 56.0 (#275)
| Benchmark | Gemma 7B | Llama 3.2 3B |
|---|---|---|
| LMArena Instruction Following | 1017 | 1089 |
Long Context Llama 3.2 3B leads
Gemma 7B: 31.1 (#282), Llama 3.2 3B: 33.4 (#261)
| Benchmark | Gemma 7B | Llama 3.2 3B |
|---|---|---|
| LMArena Longer Query | 1022 | 1100 |
Writing & Preference Gemma 7B leads
Gemma 7B: 27.1 (#302), Llama 3.2 3B: 24.7 (#307)
| Benchmark | Gemma 7B | Llama 3.2 3B |
|---|---|---|
| LMArena Text | 1056 | 1110 |
| LMArena Creative Writing | 1024 | 1094 |
| LMArena Multi-Turn | 963 | 1105 |
| EQ-Bench Creative Writing | — | 595 |
Frequently asked questions
Is Gemma 7B better than Llama 3.2 3B?
Gemma 7B is the stronger model overall, scoring 30.0 to 28.9 on the Noometry Index.
Is Gemma 7B or Llama 3.2 3B better for coding?
Gemma 7B scores higher on coding benchmarks: 30.5 versus 27.6 in the Noometry coding category.
How many benchmarks do Gemma 7B and Llama 3.2 3B share?
12 benchmarks have published results for both models. Gemma 7B has 27 scored results on Noometry and Llama 3.2 3B has 18.