Qwen: Qwen3 VL 8B Thinking

qwen/qwen3-vl-8b-thinking

Qwen3-VL-8B-Thinking is the reasoning-optimized variant of the Qwen3-VL-8B multimodal model, designed for advanced visual and textual reasoning across complex scenes, documents, and temporal sequences. It integrates enhanced multimodal alignment and...

Modalities
ImageText
In / out price
₩316 / ₩3,686
1M
Context
131.07K
Released
Oct 15, 2025

Pricing

Prices are quoted in Korean won. Changes to the exchange rate or margin show up immediately.

Input (prompt)₩316 / 1M
Output (completion)₩3,686 / 1M
Cache read₩32 / 1M
Cache write

1,000 input tokens plus 1,000 output tokens costs about ₩4.