Model comparison
Gemma 3 27B vs Qwen2.5-Coder-32B
Qwen2.5-Coder-32B is the stronger model overall, scoring 32.9 to 30.5 on the Noometry Index. Gemma 3 27B costs 7.4× less per token, which makes it the better buy when Qwen2.5-Coder-32B's lead doesn't matter for your workload.
Last verified . 21 shared benchmarks.
Google
Alibaba (Qwen)Summary
- They share 21 benchmarks with published results for both. Gemma 3 27B scores higher in 3 categories and Qwen2.5-Coder-32B in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Gemma 3 27B leads 52.7 to 41.8.
- The biggest single-benchmark swing is LiveBench Coding: 39.9% for Gemma 3 27B and 56.9% for Qwen2.5-Coder-32B.
- Gemma 3 27B is cheaper at $0.08 / $0.16 per million input/output tokens, against $0.66 / $1 for Qwen2.5-Coder-32B.
- Gemma 3 27B accepts more context: 131K tokens versus 33K.
Side by side
| Gemma 3 27B | Qwen2.5-Coder-32B | |
|---|---|---|
| Provider | Alibaba (Qwen) | |
| Noometry Index | 30.5 | 32.9 |
| Released | 2025-03-11 | 2024-09-18 |
| Weights | Open | Open |
| Context window | 131K | 33K |
| Max output | 8K | 29K |
| Input $ / M tokens | $0.08 | $0.66 |
| Output $ / M tokens | $0.16 | $1 |
| Results tracked | 43 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Gemma 3 27B: 22.1 (#329), Qwen2.5-Coder-32B: 22.4 (#328)
| Benchmark | Gemma 3 27B | Qwen2.5-Coder-32B |
|---|---|---|
| Aider Polyglot | 4.9% | 16.4% |
| LiveBench Coding | 39.9% | 56.9% |
| LMArena Coding | 1322 | 1276 |
| SWE-bench Verified (bash only) | — | 9% |
| SciCode | 21.2% | — |
| BigCodeBench Instruct | — | 49% |
| BigCodeBench Complete | — | 58% |
| HumanEval+ | — | 87.2% |
| MBPP+ | — | 77% |
Agentic & Tool Use Not comparable
Gemma 3 27B: 24.9 (#105), Qwen2.5-Coder-32B: —
| Benchmark | Gemma 3 27B | Qwen2.5-Coder-32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | — |
Reasoning Qwen2.5-Coder-32B leads
Gemma 3 27B: 16.3 (#297), Qwen2.5-Coder-32B: 20.4 (#220)
| Benchmark | Gemma 3 27B | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Reasoning | 43.8% | 42.1% |
| LMArena Hard Prompts | 1340 | 1251 |
| LiveBench Data Analysis | 51.5% | 49.9% |
| Epoch Capabilities Index | 130.04 | 119.41 |
| LiveBench | 50% | 46.2% |
| Kagi LLM Benchmark | 40.4% | — |
| CritPt | 0% | — |
| Chess Puzzles | 0% | — |
| DTBench | 52.5% | — |
| LMCA | 12.3% | — |
| HellaSwag | — | 83% |
| WinoGrande | — | 80.8% |
Math Qwen2.5-Coder-32B leads
Gemma 3 27B: 24.4 (#261), Qwen2.5-Coder-32B: 30.8 (#203)
| Benchmark | Gemma 3 27B | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Math | 55.4% | 46.6% |
| LMArena Math | 1312 | 1251 |
| OTIS Mock AIME 2024-2025 | 22.5% | — |
| MATH Level 5 | 74% | — |
| GSM8K | — | 93% |
Knowledge Qwen2.5-Coder-32B leads
Gemma 3 27B: 25.4 (#256), Qwen2.5-Coder-32B: 32.9 (#200)
| Benchmark | Gemma 3 27B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Expert | 1303 | 1221 |
| GPQA Diamond | 47.7% | — |
| Confabulations | 40.3% | — |
| Vectara Hallucination Rate | 7.4% | — |
| ARC (AI2) Challenge | — | 70.5% |
| MMLU | — | 79.1% |
Multimodal Not comparable
Gemma 3 27B: 32.0 (#94), Qwen2.5-Coder-32B: —
| Benchmark | Gemma 3 27B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Vision | 1165 | — |
| GeoBench | 52% | — |
Multilingual Gemma 3 27B leads
Gemma 3 27B: 47.0 (#150), Qwen2.5-Coder-32B: 38.0 (#230)
| Benchmark | Gemma 3 27B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Non-English | 1334 | 1205 |
| LMArena Chinese | 1347 | 1222 |
| LMArena Russian | 1349 | 1228 |
| LMArena French | 1368 | — |
| LMArena German | 1362 | — |
| LMArena Japanese | 1287 | — |
| LMArena Korean | 1308 | — |
| LMArena Spanish | 1348 | — |
Instruction Following Gemma 3 27B leads
Gemma 3 27B: 70.8 (#155), Qwen2.5-Coder-32B: 61.6 (#240)
| Benchmark | Gemma 3 27B | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Instruction Following | 74.9% | 58.7% |
| LMArena Instruction Following | 1321 | 1223 |
Long Context Qwen2.5-Coder-32B leads
Gemma 3 27B: 27.6 (#288), Qwen2.5-Coder-32B: 37.9 (#204)
| Benchmark | Gemma 3 27B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Longer Query | 1333 | 1251 |
| Fiction.LiveBench | 33.3% | — |
Writing & Preference Gemma 3 27B leads
Gemma 3 27B: 52.7 (#161), Qwen2.5-Coder-32B: 41.8 (#235)
| Benchmark | Gemma 3 27B | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Text | 1358 | 1230 |
| LMArena Creative Writing | 1346 | 1174 |
| LMArena Multi-Turn | 1344 | 1222 |
| LiveBench Language | 34.6% | 23.3% |
| Short-Story Creative Writing | 79.9% | — |
| EQ-Bench Creative Writing | 1266 | — |
Frequently asked questions
Is Gemma 3 27B better than Qwen2.5-Coder-32B?
Qwen2.5-Coder-32B is the stronger model overall, scoring 32.9 to 30.5 on the Noometry Index. Gemma 3 27B costs 7.4× 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 27B or Qwen2.5-Coder-32B?
Gemma 3 27B is cheaper. It lists at $0.08 per million input tokens and $0.16 per million output tokens; Qwen2.5-Coder-32B lists at $0.66 and $1.
Is Gemma 3 27B or Qwen2.5-Coder-32B better for coding?
They score almost the same on coding (22.1 vs 22.4); test both on your own repository before choosing.
Which has the bigger context window?
Gemma 3 27B does, with 131K tokens against 33K.
How many benchmarks do Gemma 3 27B and Qwen2.5-Coder-32B share?
21 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and Qwen2.5-Coder-32B has 31.