Qwen API pricing
Qwen input, cached-input and output prices on one key, read live from the catalog. Prepaid by card or crypto, no KYC, no subscription.
Qwen3.7 Max for the hardest prompts, a Flash or Plus id for volume, the open-weight 27B and 35B ids when you want a smaller model with the same behaviour you run locally. Every row below is a live catalog read; prices are USD per million tokens and the developer list appears where recorded.
01Every Qwen model
| Model | Input | Cached | Output | Developer list in / out | Context |
|---|---|---|---|---|---|
| Qwen3.7 Max | $0.6471 (56% under) | $0.0885 | $1.9412 (56% under) | $1.475 / $4.425 | 1M tokens |
| Qwen3.6 27B | $0.09 (70% under) | input rate | $0.60 (70% under) | $0.30 / $2.00 | 262K tokens |
| Qwen3.8 Flash | $0.1152 (23% under) | $0.0123 | $0.361 (23% under) | $0.15 / $0.47 | 992K tokens |
| Qwen3.8 Max | $0.80 (60% under) | $0.60 | $2.40 (60% under) | $2.00 / $6.00 | 991K tokens |
| Qwen Flash | $0.12 | $0.0704 | $0.96 | — | — |
| Qwen MT Plus | $1.00 | $1.00 | $3.00 | — | — |
| Qwen2.5 7B Instruct | $0.20 | $0.20 | $0.20 | — | — |
| Qwen3 30B-A3B Instruct 2507 | $0.50 | input rate | $1.25 | — | — |
| Qwen3 32B | $0.56 | input rate | $2.28 | — | — |
| Qwen3 Coder 30B-A3B Instruct | $0.28 | input rate | $1.12 | — | — |
| Qwen3 Coder 480B-A35B Instruct | $0.52 | input rate | $2.00 | — | 262K tokens |
| Qwen3 Coder Flash | $0.78 | $0.156 | $3.90 | — | — |
| Qwen3 Embedding 0.6B | $0.0295 | input rate | $0.00 | — | 8K tokens |
| Qwen3 Embedding 8B | $0.0313 | input rate | $0.00 | — | 33K tokens |
| Qwen3.5 122B-A10B | $1.48 | input rate | $11.76 | — | — |
| Qwen3.5 27B | $1.12 | input rate | $8.80 | — | — |
| Qwen3.5 35B-A3B | $0.72 | $0.44 | $1.84 | — | — |
| Qwen3.5 9B | $0.40 | input rate | $0.60 | — | — |
| Qwen3.5 Flash | $0.36 | $0.08 | $1.48 | — | — |
| Qwen3.6 35B-A3B | $1.08 | $0.48 | $6.28 | — | — |
| Qwen3.8 27B | $0.48 | $0.10 | $3.20 | — | — |
USD per million tokens, read from the RelaySwitch catalog when this page was built. Developer list: the catalog's record of the developer's own price.
02Which model should I use?
Qwen3.7 Max for the hardest prompts, a Flash or Plus id for volume, the open-weight 27B and 35B ids when you want a smaller model with the same behaviour you run locally. Compare the worked models below before choosing for production.
03Worked models
- Qwen3.7 Max — 0.6471 input / 1.9412 output per million tokens.
- Qwen3.6 27B — 0.0900 input / 0.6000 output per million tokens.
Questions
- Is Qwen billed monthly?
- No. Credit is prepaid and each request bills per token.
- Can I pay for Qwen with crypto?
- Yes: USDT, USDC, BTC, ETH or SOL, alongside card.