Model comparison
Gemma 3 12B vs Qwen2-72B
Gemma 3 12B is the stronger model overall, scoring 32.1 to 30.0 on the Noometry Index.
Last verified . 14 shared benchmarks.
Google
Alibaba (Qwen)Summary
- They share 14 benchmarks with published results for both. Gemma 3 12B scores higher in 7 categories and Qwen2-72B in 2 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in multilingual, where Gemma 3 12B leads 45.7 to 35.9.
Side by side
| Gemma 3 12B | Qwen2-72B | |
|---|---|---|
| Provider | Alibaba (Qwen) | |
| Noometry Index | 32.1 | 30.0 |
| Released | 2025-03-12 | 2024-06-07 |
| Weights | Open | Open |
| Context window | 131K | — |
| Max output | 8K | — |
| Input $ / M tokens | $0.05 | — |
| Output $ / M tokens | $0.15 | — |
| Results tracked | 24 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 3 12B leads
Gemma 3 12B: 31.7 (#280), Qwen2-72B: 29.1 (#310)
| Benchmark | Gemma 3 12B | Qwen2-72B |
|---|---|---|
| LMArena Coding | 1281 | 1196 |
| SciCode | 17.4% | — |
| WeirdML | — | 11.3% |
| BigCodeBench Instruct | — | 38.5% |
| BigCodeBench Complete | — | 54% |
Agentic & Tool Use Gemma 3 12B leads
Gemma 3 12B: 25.5 (#108), Qwen2-72B: 17.0 (#146)
| Benchmark | Gemma 3 12B | Qwen2-72B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 30.4% | — |
| TheAgentCompany | — | 1.1% |
| METR Time Horizons | — | 29.9% |
Reasoning Qwen2-72B leads
Gemma 3 12B: 15.7 (#313), Qwen2-72B: 23.2 (#181)
| Benchmark | Gemma 3 12B | Qwen2-72B |
|---|---|---|
| LMArena Hard Prompts | 1309 | 1191 |
| Epoch Capabilities Index | 123.5 | 125.28 |
| CritPt | 0% | — |
| Chess Puzzles | 0% | — |
| DTBench | 48.8% | — |
| LMCA | 4.5% | — |
Math Qwen2-72B leads
Gemma 3 12B: 22.3 (#279), Qwen2-72B: 30.2 (#236)
| Benchmark | Gemma 3 12B | Qwen2-72B |
|---|---|---|
| LMArena Math | 1307 | 1235 |
| OTIS Mock AIME 2024-2025 | 16.7% | — |
| MATH Level 5 | — | 39.1% |
Knowledge Gemma 3 12B leads
Gemma 3 12B: 26.5 (#257), Qwen2-72B: 21.2 (#275)
| Benchmark | Gemma 3 12B | Qwen2-72B |
|---|---|---|
| GPQA Diamond | 39.5% | 40.8% |
| LMArena Expert | 1248 | 1171 |
| Vectara Hallucination Rate | 4.4% | — |
| MMLU | — | 82.4% |
Multimodal Not comparable
Gemma 3 12B: —, Qwen2-72B: —
| Benchmark | Gemma 3 12B | Qwen2-72B |
|---|---|---|
| MindCube | 46.7% | — |
Multilingual Gemma 3 12B leads
Gemma 3 12B: 45.7 (#165), Qwen2-72B: 35.9 (#244)
| Benchmark | Gemma 3 12B | Qwen2-72B |
|---|---|---|
| LMArena Non-English | 1318 | 1176 |
| LMArena German | 1370 | 1151 |
| LMArena Russian | 1335 | 1169 |
| LMArena Chinese | — | 1240 |
| LMArena French | — | 1170 |
| LMArena Japanese | — | 1111 |
| LMArena Korean | — | 1083 |
| LMArena Spanish | — | 1169 |
Instruction Following Gemma 3 12B leads
Gemma 3 12B: 68.6 (#186), Qwen2-72B: 61.7 (#241)
| Benchmark | Gemma 3 12B | Qwen2-72B |
|---|---|---|
| LMArena Instruction Following | 1299 | 1181 |
Long Context Gemma 3 12B leads
Gemma 3 12B: 40.0 (#162), Qwen2-72B: 36.1 (#235)
| Benchmark | Gemma 3 12B | Qwen2-72B |
|---|---|---|
| LMArena Longer Query | 1317 | 1192 |
Writing & Preference Gemma 3 12B leads
Gemma 3 12B: 47.5 (#209), Qwen2-72B: 40.8 (#241)
| Benchmark | Gemma 3 12B | Qwen2-72B |
|---|---|---|
| LMArena Text | 1334 | 1203 |
| LMArena Creative Writing | 1331 | 1181 |
| LMArena Multi-Turn | 1334 | 1196 |
| EQ-Bench Creative Writing | 1126 | — |
Frequently asked questions
Is Gemma 3 12B better than Qwen2-72B?
Gemma 3 12B is the stronger model overall, scoring 32.1 to 30.0 on the Noometry Index.
Is Gemma 3 12B or Qwen2-72B better for coding?
Gemma 3 12B scores higher on coding benchmarks: 31.7 versus 29.1 in the Noometry coding category.
How many benchmarks do Gemma 3 12B and Qwen2-72B share?
14 benchmarks have published results for both models. Gemma 3 12B has 24 scored results on Noometry and Qwen2-72B has 26.