Model comparison
Gemma 3 12B vs Qwen2.5-Coder-32B
Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 32.1 on the Noometry Index. Gemma 3 12B costs 9.9× less per token, which makes it the better buy when Qwen2.5-Coder-32B's lead doesn't matter for your workload.
Last verified . 12 shared benchmarks.
Google
Alibaba (Qwen)Summary
- They share 12 benchmarks with published results for both. Gemma 3 12B scores higher in 5 categories and Qwen2.5-Coder-32B in 3 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen2.5-Coder-32B leads 33.3 to 22.3.
- Gemma 3 12B is cheaper at $0.05 / $0.15 per million input/output tokens, against $0.66 / $1 for Qwen2.5-Coder-32B.
- Gemma 3 12B accepts more context: 131K tokens versus 33K.
Side by side
| Gemma 3 12B | Qwen2.5-Coder-32B | |
|---|---|---|
| Provider | Alibaba (Qwen) | |
| Noometry Index | 32.1 | 33.4 |
| Released | 2025-03-12 | 2024-09-18 |
| Weights | Open | Open |
| Context window | 131K | 33K |
| Max output | 8K | 29K |
| Input $ / M tokens | $0.05 | $0.66 |
| Output $ / M tokens | $0.15 | $1 |
| Results tracked | 24 | 31 |
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), Qwen2.5-Coder-32B: 22.6 (#333)
| Benchmark | Gemma 3 12B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Coding | 1281 | 1276 |
| SWE-bench Verified (bash only) | — | 9% |
| Aider Polyglot | — | 16.4% |
| SciCode | 17.4% | — |
| BigCodeBench Instruct | — | 49% |
| LiveBench Coding | — | 56.9% |
| BigCodeBench Complete | — | 58% |
| HumanEval+ | — | 87.2% |
| MBPP+ | — | 77% |
Agentic & Tool Use Not comparable
Gemma 3 12B: 25.5 (#108), Qwen2.5-Coder-32B: —
| Benchmark | Gemma 3 12B | Qwen2.5-Coder-32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 30.4% | — |
Reasoning Qwen2.5-Coder-32B leads
Gemma 3 12B: 15.7 (#313), Qwen2.5-Coder-32B: 21.2 (#225)
| Benchmark | Gemma 3 12B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Hard Prompts | 1309 | 1251 |
| Epoch Capabilities Index | 123.5 | 119.49 |
| CritPt | 0% | — |
| Chess Puzzles | 0% | — |
| LiveBench Reasoning | — | 42.1% |
| DTBench | 48.8% | — |
| LiveBench Data Analysis | — | 49.9% |
| LMCA | 4.5% | — |
| HellaSwag | — | 83% |
| LiveBench | — | 46.2% |
| WinoGrande | — | 80.8% |
Math Qwen2.5-Coder-32B leads
Gemma 3 12B: 22.3 (#279), Qwen2.5-Coder-32B: 33.3 (#204)
| Benchmark | Gemma 3 12B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Math | 1307 | 1251 |
| OTIS Mock AIME 2024-2025 | 16.7% | — |
| LiveBench Math | — | 46.6% |
| GSM8K | — | 93% |
Knowledge Qwen2.5-Coder-32B leads
Gemma 3 12B: 26.5 (#257), Qwen2.5-Coder-32B: 33.4 (#203)
| Benchmark | Gemma 3 12B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Expert | 1248 | 1221 |
| GPQA Diamond | 39.5% | — |
| Vectara Hallucination Rate | 4.4% | — |
| ARC (AI2) Challenge | — | 70.5% |
| MMLU | — | 79.1% |
Multimodal Not comparable
Gemma 3 12B: —, Qwen2.5-Coder-32B: —
| Benchmark | Gemma 3 12B | Qwen2.5-Coder-32B |
|---|---|---|
| MindCube | 46.7% | — |
Multilingual Gemma 3 12B leads
Gemma 3 12B: 45.7 (#165), Qwen2.5-Coder-32B: 37.8 (#235)
| Benchmark | Gemma 3 12B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Non-English | 1318 | 1205 |
| LMArena Russian | 1335 | 1228 |
| LMArena Chinese | — | 1222 |
| LMArena German | 1370 | — |
Instruction Following Gemma 3 12B leads
Gemma 3 12B: 68.6 (#186), Qwen2.5-Coder-32B: 61.4 (#245)
| Benchmark | Gemma 3 12B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Instruction Following | 1299 | 1223 |
| LiveBench Instruction Following | — | 58.7% |
Long Context Gemma 3 12B leads
Gemma 3 12B: 40.0 (#162), Qwen2.5-Coder-32B: 38.0 (#208)
| Benchmark | Gemma 3 12B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Longer Query | 1317 | 1251 |
Writing & Preference Gemma 3 12B leads
Gemma 3 12B: 47.5 (#209), Qwen2.5-Coder-32B: 41.6 (#240)
| Benchmark | Gemma 3 12B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Text | 1334 | 1230 |
| LMArena Creative Writing | 1331 | 1174 |
| LMArena Multi-Turn | 1334 | 1222 |
| EQ-Bench Creative Writing | 1126 | — |
| LiveBench Language | — | 23.3% |
Frequently asked questions
Is Gemma 3 12B better than Qwen2.5-Coder-32B?
Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 32.1 on the Noometry Index. Gemma 3 12B costs 9.9× less per token, which makes it the better buy when Qwen2.5-Coder-32B's lead doesn't matter for your workload.
Which is cheaper, Gemma 3 12B or Qwen2.5-Coder-32B?
Gemma 3 12B is cheaper. It lists at $0.05 per million input tokens and $0.15 per million output tokens; Qwen2.5-Coder-32B lists at $0.66 and $1.
Is Gemma 3 12B or Qwen2.5-Coder-32B better for coding?
Gemma 3 12B scores higher on coding benchmarks: 31.7 versus 22.6 in the Noometry coding category.
Which has the bigger context window?
Gemma 3 12B does, with 131K tokens against 33K.
How many benchmarks do Gemma 3 12B and Qwen2.5-Coder-32B share?
12 benchmarks have published results for both models. Gemma 3 12B has 24 scored results on Noometry and Qwen2.5-Coder-32B has 31.