Qwen3 235B A22B Thinking 2507 is cheaper than Grok Code Fast 1 at $0.11/M vs $0.2/M input tokens.
| Spec | Grok Code Fast 1 | Qwen3 235B A22B Thinking 2507 |
|---|---|---|
| Input price | $0.2/M tokens | $0.11/M tokens |
| Output price | $1.5/M tokens | $0.6/M tokens |
| Context window | 256K tokens | 131K tokens |
| Weights | — | Open |
| Free API (OpenRouter) | No | No |
| Released | Aug 2025 | Jul 2025 |
Not enough votes to call it. On the specs, nothing separates them.
Qwen3 235B A22B Thinking 2507 is cheaper on both: 1.8× input, 2.5× output.
No hosts listed on OpenRouter.
Per million tokens. Prices and uptime via OpenRouter, checked 8 Sep 2026.
Qwen3 235B A22B Thinking 2507 uses 21.0x more emoji