Qwen3.5-Flash
Qwen · qwen/qwen3.5-flash-02-23
The Qwen3.5 native vision-language Flash models are built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. Compared to the...
Qwen3.5-Flash ranks #52 of 377 models in the default AI Benchmark Hub order (quality first, then cost). At $0.16 blended per million tokens, it is cheaper than 86% of models with published API pricing. Its 1,000,000-token context window is 3.8× the catalog median (262,144 tokens).
Context
Max output: 65536
Pricing
Output / 1M: 0.26
Blend / 1M: 0.16
Quality
Provider
Moderated: no
Monthly cost at this blended rate
Estimated API spend if every token is billed at $0.16per million (the blend of this model's input and output prices). Real bills depend on the input/output mix.
| Tokens / month | Est. cost (USD) |
|---|---|
| 1,000,000 | $0.16 |
| 10,000,000 | $1.63 |
| 100,000,000 | $16.25 |
Related models
Same lab, similar price, or similar quality — useful next comparisons.
More from Qwen
- Qwen3.7 Flash$0.08 / 1M
- Qwen2.5 7B Instruct$0.15 / 1M
- Qwen3 14B$0.18 / 1M
- Qwen3 32B$0.18 / 1M
- Qwen3 30B A3B Instruct 2507$0.20 / 1M
Closest API price
- Nex-N2.5-Pro$0.16 / 1M
- Granite 4.2 8B$0.15 / 1M
- DeepSeek V4 Flash Latest$0.17 / 1M
- DeepSeek V4 Flash 0731$0.17 / 1M
- Ling 3.1 Flash$0.15 / 1M