Compare / Qwen3.6 27B vs DiffusionGemma 26B-A4B
Qwen3.6 27B vs DiffusionGemma 26B-A4B
Updated November 2026
Verdict
Qwen3.6 27B scores higher on the Small LM Index: 74.0 vs 52.6. On the 5 benchmarks both labs report, Qwen3.6 27B wins 5 and DiffusionGemma 26B-A4B wins 0. Both need about 17.9 GB of memory at 4-bit quantization. Only Qwen3.6 27B has a hosted API on OpenRouter ($0.90 per million tokens); the other has to be self-hosted.
Pick Qwen3.6 27B if…
- you want the strongest model overall
- you mostly write code
- you need math and reasoning
- you want a hosted API rather than running it yourself
- you work with long documents
- you want the more widely used model (more quantized builds and community support)
Pick DiffusionGemma 26B-A4B if…
No clear advantage in the data we have.
Specifications
| Qwen3.6 27B | DiffusionGemma 26B-A4B | |
|---|---|---|
| Small LM Index | 74.0 (likely 73–75) | 52.6 (likely 49–56) |
| Lab | Alibaba (Qwen) · China | Google · United States |
| Released | 2026-04-22 | 2026-06-09 |
| Parameters | 27B (dense) | 25.8B (4B active, MoE) |
| Memory at 4-bit | ~17.9 GB | ~17.2 GB |
| Context window | 262K tokens | 256K tokens |
| Input | text, image, video | text, image |
| Reasoning | Hybrid (thinking on/off) | Hybrid (thinking on/off) |
| License | Apache-2.0 | Apache-2.0 |
| API price / 1M tokens | $0.90 ($0.20 in, $3.00 out) | Self-host only |
| Hugging Face downloads (30 days) | 2.3M | 625.6K |
Benchmarks
Higher is better. Bold marks the winner where both labs report a score. ≈ marks the index model's estimate for a score that wasn't reported, ✓ an independent measurement and * a score measured by another lab.
| Benchmark | Qwen3.6 27B | DiffusionGemma 26B-A4B |
|---|---|---|
| MMLU-Pro Knowledge | 86.2 | 77.6 |
| GPQA Diamond Knowledge | 87.8 | 73.2 |
| Humanity's Last Exam Reasoning | 24.0 | 11.0 |
| AIME Math | 94.1 | 69.1 |
| LiveCodeBench Coding | 83.9 | 69.1 |
| SWE-bench Verified Coding | 77.2 | ≈ 35 |
| Terminal-Bench Agentic | 59.3 | — |
| τ²-Bench Agentic | — | 56.2 |
| MMMU-Pro Vision | — | 54.3 |
| Multilingual MMLU Multilingual | — | 81.5 |
Scores are mostly reported by each lab with its own settings, so small gaps are noise. See the methodology. Open in the interactive comparison to add more models.
Questions
Which is better, Qwen3.6 27B or DiffusionGemma 26B-A4B?
Qwen3.6 27B scores higher on the Small LM Index: 74.0 vs 52.6. On the 5 benchmarks both labs report, Qwen3.6 27B wins 5 and DiffusionGemma 26B-A4B wins 0.
Which is cheaper to run?
Only Qwen3.6 27B has a hosted API on OpenRouter ($0.90 per million tokens); the other has to be self-hosted. Both need about 17.9 GB of memory at 4-bit quantization.
Which has the longer context window?
Qwen3.6 27B has the longer context window: 262K tokens against 256K.
Can I use Qwen3.6 27B and DiffusionGemma 26B-A4B commercially?
Both are released under the Apache-2.0 license. Apache 2.0 and MIT allow commercial use; other licenses can add conditions, so read the license before shipping a product.