Qwen3.8 Flash
by Alibaba (China) · family qwen · listed Aug 2026RTV
Qwen vision-language model for visual reasoning, documents, and agent tasks
Current rates USD per 1M tokens
Input$0.119
Output$0.401
Cache read$0.0119
Cache write$0.148
Against the market
On the record
- Model id
qwen3.8-flash- Context window
- 1M tokens
- Max output
- 131K tokens
- Input modalities
- text, image, video
- Output modalities
- text
- Reasoning
- yes
- Tool calls
- yes
- Open weights
- no
- Knowledge cutoff
- —
- Released
- 2026-08-26
- Last updated
- 2026-08-26
- Provider docs
- www.alibabacloud.com/help/en/model-studio/models
Back of the envelope
≈
Also listed at
- Alibaba Token Planfree in · free out
- Alibaba Token Plan (China)free in · free out
- Vancine$0.12 in · $0.38 out
- CrossModel$0.13 in · $0.43 out
- OpenRouter$0.15 in · $0.47 out
- OpenCode Go$0.15 in · $0.47 out
- LLM Gateway$0.15 in · $0.47 out
- Alibaba$0.15 in · $0.47 out
- Charm Hyper$0.15 in · $0.47 out
- DevPass (LLM Gateway)$0.15 in · $0.47 out
- Kilo Gateway$0.15 in · $0.47 out
- NanoGPT$0.16 in · $0.47 out
- Vercel AI Gateway$0.16 in · $0.47 out
- Eden AI$0.16 in · $0.47 out