Qwen 3 235 B A22 B2507 pricing
Source: openrouter · updated 2026-10-27
Short answer: Qwen 3 235 B A22 B2507 costs $0.090 per million input tokens and $0.550 per million output tokens. That is a 3:1 blended rate of $0.205 per million tokens, ranking it #114 of 704 tracked models by input price.
Price breakdown
| Metric | USD |
|---|---|
| Input, per 1M tokens | $0.090 |
| Output, per 1M tokens | $0.550 |
| Output รท input multiple | 6.11× |
| 3:1 blended, per 1M tokens | $0.205 |
| Context window | 262,144 tokens |
What a typical workload costs
| Monthly volume | Input cost | Output cost | Total | 1M in / 250K out | $0.090 | $0.138 | $0.228 |
|---|---|---|---|---|---|---|---|
| 10M in / 2.5M out | $0.900 | $1.38 | $2.27 | ||||
| 100M in / 25M out | $9.00 | $13.75 | $22.75 |
Other qwen models
| Model | Input | Output |
|---|---|---|
| Qwen 3.7 Flash | $0.030 | $0.130 |
| Qwen 3.5 Flash 02 23 | $0.065 | $0.260 |
| Qwen 3 Coder 30 B A3 B Instruct | $0.070 | $0.280 |
| Qwen 3 32 B | $0.080 | $0.280 |
| Qwen 3 Next 80 B A3 B Instruct | $0.090 | $1.10 |
Compare against
- Qwen 3 235 B A22 B2507 vs Mistral Small 3.2 24 B Instruct 2506
- Qwen 3 235 B A22 B2507 vs Qwen 3 Next 80 B A3 B Instruct
Cheaper alternatives
Next cheaper option: GPT Luna Latest at $0.100 per 1M input tokens. Next step up: Llama 3.2 11 B Vision Instruct at $0.049 per 1M input tokens.