Qwen3 235B A22B Thinking 2507
Alibaba · Released Jul 2025
A 235-billion-parameter mixture-of-experts model from Alibaba that activates 22 billion parameters per token and uses extended reasoning capabilities.
- Strengths
- Handles complex problems through step-by-step reasoning, making it effective for tasks requiring deliberate analysis and multi-step logic.
- Best for
- Workloads that benefit from reasoning overhead — mathematical problem-solving, code debugging, structured analysis, and other tasks where tracing through the reasoning process matters.
- Limitations
- Reasoning adds latency and token consumption compared to the standard Instruct variant, so it trades speed and efficiency for reasoning depth.
Input / 1M
$0.23
Output / 1M
$2.30
Cached input / 1M
—
Context window
131K
Price history
Input (solid)Output (dashed)
Price change
- 30d
- in — out —
- 90d
- in increased 130.0% out increased 2200.0%
- 1y
- in increased 130.0% out increased 2200.0%
- Since launch
- in increased 130.0% out increased 2200.0%
Snapshots
| Effective | Input | Output | Cached in | Note | Source |
|---|---|---|---|---|---|
| 31 Jul 2026 | $0.23 | $2.30 | — | Imported from OpenRouter | openrouter.ai |
| 20 Jul 2026 | $0.3 | $3.00 | — | Imported from OpenRouter | openrouter.ai |
| 1 Jul 2026 | $0.1495 | $1.50 | — | Imported from OpenRouter | openrouter.ai |
| 11 Jun 2026 | $0.1 | $0.1 | $0.1 | Imported from OpenRouter | openrouter.ai |