Model comparison
Gemma 3 4B vs Llama 2-13B
Llama 2-13B is the stronger model overall, scoring 29.6 to 28.1 on the Noometry Index.
Last verified . 15 shared benchmarks.
Google
MetaSummary
- They share 15 benchmarks with published results for both. Gemma 3 4B scores higher in 6 categories and Llama 2-13B in 2 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Llama 2-13B leads 28.1 to 11.8.
- The biggest single-benchmark swing is DTBench: 50.9% for Gemma 3 4B and 42.2% for Llama 2-13B.
Side by side
| Gemma 3 4B | Llama 2-13B | |
|---|---|---|
| Provider | Meta | |
| Noometry Index | 28.1 | 29.6 |
| Released | 2025-03-12 | 2023-07-18 |
| Weights | Open | Open |
| Context window | 131K | — |
| Max output | 4K | — |
| Input $ / M tokens | $0.04 | — |
| Output $ / M tokens | $0.08 | — |
| Results tracked | 22 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 3 4B leads
Gemma 3 4B: 35.9 (#215), Llama 2-13B: 30.9 (#291)
| Benchmark | Gemma 3 4B | Llama 2-13B |
|---|---|---|
| LMArena Coding | 1230 | 1062 |
Agentic & Tool Use Not comparable
Gemma 3 4B: 20.9 (#142), Llama 2-13B: —
| Benchmark | Gemma 3 4B | Llama 2-13B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 19.6% | — |
Reasoning Too close to call
Gemma 3 4B: 13.2 (#335), Llama 2-13B: 12.8 (#337)
| Benchmark | Gemma 3 4B | Llama 2-13B |
|---|---|---|
| Chess Puzzles | 0% | 0% |
| LMArena Hard Prompts | 1253 | 1051 |
| DTBench | 50.9% | 42.2% |
| Epoch Capabilities Index | 116.02 | 106.17 |
| Kagi LLM Benchmark | 25.2% | — |
| LMCA | 2.8% | — |
| BIG-Bench Hard | — | 58.2% |
| HellaSwag | — | 80.7% |
| LAMBADA | — | 76.5% |
| PIQA | — | 80.8% |
| WinoGrande | — | 72.8% |
Math Llama 2-13B leads
Gemma 3 4B: 16.8 (#292), Llama 2-13B: 31.1 (#229)
| Benchmark | Gemma 3 4B | Llama 2-13B |
|---|---|---|
| LMArena Math | 1239 | 1065 |
| OTIS Mock AIME 2024-2025 | 7.5% | — |
| GSM8K | — | 36.9% |
Knowledge Llama 2-13B leads
Gemma 3 4B: 11.8 (#299), Llama 2-13B: 28.1 (#249)
| Benchmark | Gemma 3 4B | Llama 2-13B |
|---|---|---|
| LMArena Expert | 1223 | 1030 |
| GPQA Diamond | 23.2% | — |
| Vectara Hallucination Rate | 6.4% | — |
| ARC (AI2) Challenge | — | 60.3% |
| BoolQ | — | 82.4% |
| MMLU | — | 55.6% |
| OpenBookQA | — | 57% |
| TriviaQA | — | 79.6% |
Multimodal Not comparable
Gemma 3 4B: —, Llama 2-13B: —
| Benchmark | Gemma 3 4B | Llama 2-13B |
|---|---|---|
| ScienceQA | — | 55.8% |
Multilingual Gemma 3 4B leads
Gemma 3 4B: 42.5 (#194), Llama 2-13B: 26.5 (#279)
| Benchmark | Gemma 3 4B | Llama 2-13B |
|---|---|---|
| LMArena Non-English | 1273 | 1024 |
| LMArena German | 1281 | 1009 |
| LMArena Russian | 1294 | 1055 |
| LMArena Chinese | — | 1001 |
| LMArena French | — | 1044 |
| LMArena Japanese | — | 894 |
| LMArena Korean | — | 953 |
| LMArena Spanish | — | 1087 |
Instruction Following Gemma 3 4B leads
Gemma 3 4B: 65.2 (#225), Llama 2-13B: 53.3 (#287)
| Benchmark | Gemma 3 4B | Llama 2-13B |
|---|---|---|
| LMArena Instruction Following | 1239 | 1045 |
Long Context Gemma 3 4B leads
Gemma 3 4B: 38.7 (#194), Llama 2-13B: 32.3 (#269)
| Benchmark | Gemma 3 4B | Llama 2-13B |
|---|---|---|
| LMArena Longer Query | 1273 | 1064 |
Writing & Preference Gemma 3 4B leads
Gemma 3 4B: 42.0 (#239), Llama 2-13B: 29.8 (#289)
| Benchmark | Gemma 3 4B | Llama 2-13B |
|---|---|---|
| LMArena Text | 1291 | 1084 |
| LMArena Creative Writing | 1271 | 1047 |
| LMArena Multi-Turn | 1255 | 1050 |
| EQ-Bench Creative Writing | 1068 | — |
Frequently asked questions
Is Gemma 3 4B better than Llama 2-13B?
Llama 2-13B is the stronger model overall, scoring 29.6 to 28.1 on the Noometry Index.
Is Gemma 3 4B or Llama 2-13B better for coding?
Gemma 3 4B scores higher on coding benchmarks: 35.9 versus 30.9 in the Noometry coding category.
How many benchmarks do Gemma 3 4B and Llama 2-13B share?
15 benchmarks have published results for both models. Gemma 3 4B has 22 scored results on Noometry and Llama 2-13B has 32.