Mistral Medium 3.5 vs Qwen3.6 27B
⚖ Mistral Medium 3.5 and Qwen3.6 27B trade wins evenly on our provisional scores, but Qwen3.6 27B is 2.2× cheaper for everyday text work. Pick Mistral Medium 3.5 when quality matters most, Qwen3.6 27B at volume.
Scores by category
| Category | Mistral Medium 3.5 | Qwen3.6 27B |
|---|---|---|
| Overall | 18 | 18 |
| Writing | 35 | – |
| Support | 40 | – |
| Coding | 19 | – |
Provisional scores, 0 to 100 among the models we track, from Arena and Epoch AI. Value weighs the overall score against cost.
Price and specs
| Mistral Medium 3.5 | Qwen3.6 27B | |
|---|---|---|
| Vendor | Mistral AI | Alibaba (Qwen) |
| Released | 2026-04-29 | 2026-04-22 |
| Context window | 262K | 262K |
| Open weights | Yes | Yes |
| Input per 1M tokens | $1.50 | $0.60 |
| Output per 1M tokens | $7.50 | $3.60 |
| Per 1,000 tasks | $6.00 | $2.70 |