Model comparison
Gemma 3 27B vs Qwen3.5 27B
Qwen3.5 27B is the stronger model overall, scoring 41.4 to 30.5 on the Noometry Index. Gemma 3 27B costs 8.2× less per token, which makes it the better buy when Qwen3.5 27B'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 0 categories and Qwen3.5 27B in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in coding, where Qwen3.5 27B leads 38.6 to 22.1.
- The biggest single-benchmark swing is DTBench: 52.5% for Gemma 3 27B and 82.4% for Qwen3.5 27B.
- Gemma 3 27B is cheaper at $0.08 / $0.16 per million input/output tokens, against $0.30 / $2.40 for Qwen3.5 27B.
- Qwen3.5 27B accepts more context: 262K tokens versus 131K.
Side by side
| Gemma 3 27B | Qwen3.5 27B | |
|---|---|---|
| Provider | Alibaba (Qwen) | |
| Noometry Index | 30.5 | 41.4 |
| Released | 2025-03-11 | 2026-02-23 |
| Weights | Open | Open |
| Context window | 131K | 262K |
| Max output | 8K | 66K |
| Input $ / M tokens | $0.08 | $0.30 |
| Output $ / M tokens | $0.16 | $2.40 |
| Results tracked | 43 | 28 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5 27B leads
Gemma 3 27B: 22.1 (#329), Qwen3.5 27B: 38.6 (#159)
| Benchmark | Gemma 3 27B | Qwen3.5 27B |
|---|---|---|
| LMArena Coding | 1322 | 1427 |
| Aider Polyglot | 4.9% | — |
| LMArena WebDev | — | 1357 |
| SciCode | 21.2% | — |
| WeirdML | — | 39.5% |
| LiveBench Coding | 39.9% | — |
| ALE-Bench | — | 349.45 |
Agentic & Tool Use Not comparable
Gemma 3 27B: 24.9 (#105), Qwen3.5 27B: —
| Benchmark | Gemma 3 27B | Qwen3.5 27B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | — |
| Vending-Bench 2 | — | 201.98 |
Reasoning Qwen3.5 27B leads
Gemma 3 27B: 16.3 (#297), Qwen3.5 27B: 26.6 (#116)
| Benchmark | Gemma 3 27B | Qwen3.5 27B |
|---|---|---|
| LMArena Hard Prompts | 1340 | 1414 |
| DTBench | 52.5% | 82.4% |
| LMCA | 12.3% | 34% |
| Kagi LLM Benchmark | 40.4% | — |
| NYT Connections (extended) | — | 47.9% |
| CritPt | 0% | — |
| Chess Puzzles | 0% | — |
| Thematic Generalization | — | 45.5% |
| LiveBench Reasoning | 43.8% | — |
| LiveBench Data Analysis | 51.5% | — |
| Epoch Capabilities Index | 130.04 | — |
| LiveBench | 50% | — |
Math Qwen3.5 27B leads
Gemma 3 27B: 24.4 (#261), Qwen3.5 27B: 36.8 (#103)
| Benchmark | Gemma 3 27B | Qwen3.5 27B |
|---|---|---|
| LMArena Math | 1312 | 1428 |
| MathArena Final-Answer Competitions | — | 56.7% |
| OTIS Mock AIME 2024-2025 | 22.5% | — |
| LiveBench Math | 55.4% | — |
| MATH Level 5 | 74% | — |
Knowledge Qwen3.5 27B leads
Gemma 3 27B: 25.4 (#256), Qwen3.5 27B: 37.4 (#149)
| Benchmark | Gemma 3 27B | Qwen3.5 27B |
|---|---|---|
| Vectara Hallucination Rate | 7.4% | 12.1% |
| LMArena Expert | 1303 | 1428 |
| GPQA Diamond | 47.7% | — |
| Confabulations | 40.3% | — |
Multimodal Qwen3.5 27B leads
Gemma 3 27B: 32.0 (#94), Qwen3.5 27B: 38.7 (#56)
| Benchmark | Gemma 3 27B | Qwen3.5 27B |
|---|---|---|
| LMArena Vision | 1165 | 1240 |
| GeoBench | 52% | — |
Multilingual Qwen3.5 27B leads
Gemma 3 27B: 47.0 (#150), Qwen3.5 27B: 51.0 (#112)
| Benchmark | Gemma 3 27B | Qwen3.5 27B |
|---|---|---|
| LMArena Non-English | 1334 | 1390 |
| LMArena Chinese | 1347 | 1478 |
| LMArena French | 1368 | 1409 |
| LMArena German | 1362 | 1393 |
| LMArena Japanese | 1287 | 1345 |
| LMArena Korean | 1308 | 1359 |
| LMArena Russian | 1349 | 1389 |
| LMArena Spanish | 1348 | 1408 |
Instruction Following Qwen3.5 27B leads
Gemma 3 27B: 70.8 (#155), Qwen3.5 27B: 73.7 (#114)
| Benchmark | Gemma 3 27B | Qwen3.5 27B |
|---|---|---|
| LMArena Instruction Following | 1321 | 1393 |
| LiveBench Instruction Following | 74.9% | — |
Long Context Qwen3.5 27B leads
Gemma 3 27B: 27.6 (#288), Qwen3.5 27B: 43.1 (#103)
| Benchmark | Gemma 3 27B | Qwen3.5 27B |
|---|---|---|
| LMArena Longer Query | 1333 | 1413 |
| Fiction.LiveBench | 33.3% | — |
Writing & Preference Qwen3.5 27B leads
Gemma 3 27B: 52.7 (#161), Qwen3.5 27B: 59.5 (#106)
| Benchmark | Gemma 3 27B | Qwen3.5 27B |
|---|---|---|
| LMArena Text | 1358 | 1409 |
| LMArena Creative Writing | 1346 | 1362 |
| LMArena Multi-Turn | 1344 | 1410 |
| Short-Story Creative Writing | 79.9% | — |
| EQ-Bench Creative Writing | 1266 | — |
| LiveBench Language | 34.6% | — |
Frequently asked questions
Is Gemma 3 27B better than Qwen3.5 27B?
Qwen3.5 27B is the stronger model overall, scoring 41.4 to 30.5 on the Noometry Index. Gemma 3 27B costs 8.2× less per token, which makes it the better buy when Qwen3.5 27B's lead doesn't matter for your workload.
Which is cheaper, Gemma 3 27B or Qwen3.5 27B?
Gemma 3 27B is cheaper. It lists at $0.08 per million input tokens and $0.16 per million output tokens; Qwen3.5 27B lists at $0.30 and $2.40.
Is Gemma 3 27B or Qwen3.5 27B better for coding?
Qwen3.5 27B scores higher on coding benchmarks: 38.6 versus 22.1 in the Noometry coding category.
Which has the bigger context window?
Qwen3.5 27B does, with 262K tokens against 131K.
How many benchmarks do Gemma 3 27B and Qwen3.5 27B share?
21 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and Qwen3.5 27B has 28.