Model comparison
Gemma 3 4B vs Gemma 7B
Gemma 7B is the stronger model overall, scoring 29.5 to 27.9 on the Noometry Index.
Last verified . 12 shared benchmarks.
GoogleSummary
- They share 12 benchmarks with published results for both. Gemma 3 4B scores higher in 5 categories and Gemma 7B in 3 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in multilingual, where Gemma 3 4B leads 42.7 to 25.3.
Side by side
| Gemma 3 4B | Gemma 7B | |
|---|---|---|
| Provider | ||
| Noometry Index | 27.9 | 29.5 |
| Released | 2025-03-12 | 2024-02-21 |
| Weights | Open | Open |
| Context window | 131K | — |
| Max output | 4K | — |
| Input $ / M tokens | $0.04 | — |
| Output $ / M tokens | $0.08 | — |
| Results tracked | 22 | 27 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 3 4B leads
Gemma 3 4B: 35.3 (#214), Gemma 7B: 29.9 (#290)
| Benchmark | Gemma 3 4B | Gemma 7B |
|---|---|---|
| LMArena Coding | 1230 | 1048 |
| HumanEval+ | — | 28.7% |
| MBPP+ | — | 43.4% |
Agentic & Tool Use Not comparable
Gemma 3 4B: 20.7 (#134), Gemma 7B: —
| Benchmark | Gemma 3 4B | Gemma 7B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 19.6% | — |
Reasoning Gemma 7B leads
Gemma 3 4B: 12.7 (#330), Gemma 7B: 19.0 (#247)
| Benchmark | Gemma 3 4B | Gemma 7B |
|---|---|---|
| LMArena Hard Prompts | 1253 | 1042 |
| Epoch Capabilities Index | 116.02 | 111.74 |
| Kagi LLM Benchmark | 25.2% | — |
| Chess Puzzles | 0% | — |
| DTBench | 50.9% | — |
| LMCA | 2.8% | — |
| Adversarial NLI | — | 48.7% |
| BIG-Bench Hard | — | 55.1% |
| HellaSwag | — | 82.2% |
| PIQA | — | 81.2% |
| WinoGrande | — | 79% |
Math Gemma 7B leads
Gemma 3 4B: 15.8 (#286), Gemma 7B: 28.7 (#225)
| Benchmark | Gemma 3 4B | Gemma 7B |
|---|---|---|
| LMArena Math | 1239 | 1066 |
| OTIS Mock AIME 2024-2025 | 7.5% | — |
| GSM8K | — | 46.4% |
Knowledge Gemma 7B leads
Gemma 3 4B: 11.8 (#294), Gemma 7B: 26.8 (#248)
| Benchmark | Gemma 3 4B | Gemma 7B |
|---|---|---|
| LMArena Expert | 1223 | 1002 |
| GPQA Diamond | 23.2% | — |
| Vectara Hallucination Rate | 6.4% | — |
| ARC (AI2) Challenge | — | 78.3% |
| BoolQ | — | 83.2% |
| MMLU | — | 66.1% |
| OpenBookQA | — | 78.6% |
| TriviaQA | — | 72.3% |
Multilingual Gemma 3 4B leads
Gemma 3 4B: 42.7 (#189), Gemma 7B: 25.3 (#282)
| Benchmark | Gemma 3 4B | Gemma 7B |
|---|---|---|
| LMArena Non-English | 1273 | 999 |
| LMArena Russian | 1294 | 993 |
| LMArena Chinese | — | 1035 |
| LMArena French | — | 1024 |
| LMArena German | 1281 | — |
Instruction Following Gemma 3 4B leads
Gemma 3 4B: 65.4 (#219), Gemma 7B: 51.7 (#290)
| Benchmark | Gemma 3 4B | Gemma 7B |
|---|---|---|
| LMArena Instruction Following | 1239 | 1017 |
Long Context Gemma 3 4B leads
Gemma 3 4B: 38.6 (#189), Gemma 7B: 31.1 (#277)
| Benchmark | Gemma 3 4B | Gemma 7B |
|---|---|---|
| LMArena Longer Query | 1274 | 1022 |
Writing & Preference Gemma 3 4B leads
Gemma 3 4B: 42.2 (#234), Gemma 7B: 27.2 (#297)
| Benchmark | Gemma 3 4B | Gemma 7B |
|---|---|---|
| LMArena Text | 1291 | 1056 |
| LMArena Creative Writing | 1270 | 1024 |
| LMArena Multi-Turn | 1255 | 963 |
| EQ-Bench Creative Writing | 1068 | — |
Frequently asked questions
Is Gemma 3 4B better than Gemma 7B?
Gemma 7B is the stronger model overall, scoring 29.5 to 27.9 on the Noometry Index.
Is Gemma 3 4B or Gemma 7B better for coding?
Gemma 3 4B scores higher on coding benchmarks: 35.3 versus 29.9 in the Noometry coding category.
How many benchmarks do Gemma 3 4B and Gemma 7B share?
12 benchmarks have published results for both models. Gemma 3 4B has 22 scored results on Noometry and Gemma 7B has 27.