Qwen3 30B A3B Thinking 2507 is cheaper than Qwen3.8 27B at $0.071/M vs $0.4/M input tokens.
| Spec | Qwen3 30B A3B Thinking 2507 | Qwen3.8 27B |
|---|---|---|
| Input price | $0.071/M tokens | $0.4/M tokens |
| Output price | $0.285/M tokens | $3/M tokens |
| Context window | 262K tokens | 262K tokens |
| Released | Aug 2025 | Aug 2026 |
Not enough votes to call it. On the specs, nothing separates them.
Qwen3 30B A3B Thinking 2507 costs 11x less per token.
Qwen3 30B A3B Thinking 2507 is cheaper on both — 5.6× input, 11× output
Qwen3 30B A3B Thinking 2507 uses 89.3x more emoji
Qwen3 30B A3B Thinking 2507 is cheaper than Qwen3.8 27B at $0.071/M vs $0.4/M input tokens.
| Spec | Qwen3 30B A3B Thinking 2507 | Qwen3.8 27B |
|---|---|---|
| Input price | $0.071/M tokens | $0.4/M tokens |
| Output price | $0.285/M tokens | $3/M tokens |
| Context window | 262K tokens | 262K tokens |
| Released | Aug 2025 | Aug 2026 |
Not enough votes to call it. On the specs, nothing separates them.
Qwen3 30B A3B Thinking 2507 costs 11x less per token.
Qwen3 30B A3B Thinking 2507 is cheaper on both — 5.6× input, 11× output
Qwen3 30B A3B Thinking 2507 uses 89.3x more emoji