Qwen3.5 27B vs Kimi K2 Thinking
Qwen3.5 27B is cheaper on output tokens. Choose Qwen3.5 27B or Kimi K2 Thinking based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Qwen3.5 27B | Kimi K2 Thinking |
|---|---|---|
| Provider | Alibaba | MoonshotAI |
| Input / 1M tokens | $0.30 | $0.60 |
| Output / 1M tokens | $2.40 | $2.50 |
| Context window | 262K | 262K |
| Parameters | — | 1T |
| Open weights | Yes | Yes |
| Released | Feb 2026 | Nov 2025 |
| LMArena (Chatbot Arena) Elo score | 1409 | — |
| LMArena WebDev Elo score | 1358 | — |
| LMArena Vision Elo score | 1241 | — |
| LiveBench score | — | 62.3 |
| Aider Polyglot score | — | 59.1 |
FAQ
Is Qwen3.5 27B or Kimi K2 Thinking cheaper?
Qwen3.5 27B is cheaper on output tokens ($2.40 vs $2.50 per 1M).
Which has the larger context window, Qwen3.5 27B or Kimi K2 Thinking?
Both models offer a similar context window.
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.