Provider · verified 16 Oct 2026
Qwen 通义千问
Qwen is the largest block of Chinese-lab ids on our OpenRouter snapshot. The flash ids are cheap on that list. The max and omni ids are not automatically cheap. Read the id before you trust a “Qwen” average.
The chat product and the API are different things. The API key does not come from the chat app.
Models on the latest OpenRouter snapshot
| Model | Input | Output | vs Opus |
|---|---|---|---|
qwen/qwen3.8-max-prime | $4.00 | $12.00 | 25% cheaper |
qwen/qwen3.8-omni-flash | $0.150 | $0.470 | 97% cheaper |
qwen/qwen3.8-max-0902 | $2.00 | $6.00 | 63% cheaper |
qwen/qwen3.8-flash | $0.150 | $0.470 | 97% cheaper |
qwen/qwen3.8-27b | $0.240 | $2.20 | 91% cheaper |
qwen/qwen3.8-2.4t-a95b | $2.00 | $6.00 | 63% cheaper |
qwen/qwen3.7-flash | $0.030 | $0.130 | 99% cheaper |
qwen/qwen3.7-plus | $0.320 | $1.28 | 93% cheaper |
qwen/qwen3.7-max | $1.48 | $4.42 | 72% cheaper |
qwen/qwen3.5-plus-20260420 | $0.300 | $1.80 | 92% cheaper |
qwen/qwen3.6-flash | $0.188 | $1.13 | 95% cheaper |
qwen/qwen3.6-35b-a3b | $0.150 | $1.00 | 95% cheaper |
Guides
- Chinese AI API prices, live, versus the latest Opus and Sol
- Call Chinese models without sending the prompt to the lab
- Use Chinese models in Claude Code: the base URLs we could verify
- Coding plans compared: GLM's credits, and the plans we would not price
- What a coding session costs: flash models, flagships, and the cache assumption
- Free Chinese AI apps: a short list that refuses fake prices
- Is it safe to use Chinese AI? A sourced risk matrix, not a slogan
- Paying for Chinese AI from abroad: what the docs show, and what they do not
- Qwen API: international sign-up, and the free tier we would not invent
- Qwen Chat: what the official app is, and what a mod apk is not
- Run Qwen locally when the API price is not the constraint
- Qwen, DeepSeek, GLM, Kimi: which Chinese model for which job