Qwen: Qwen3 235B A22B Thinking 2507

qwen/qwen3-235b-a22b-thinking-2507

Qwen3-235B-A22B-Thinking-2507 is a high-performance, open-weight Mixture-of-Experts (MoE) language model optimized for complex reasoning tasks. It activates 22B of its 235B parameters per forward pass and natively supports up to 262,144...

Modalities
Teks
Harga in / out
₩404 / ₩4.037
1M
Context
131,07 rb
Dirilis
25 Jul 2025

Pricing

Harga dinyatakan dalam KRW. Perubahan pada nilai tukar atau margin akan langsung diterapkan.

Input (prompt)₩404 / 1M
Output (completion)₩4.037 / 1M
Cache read₩40 / 1M
Cache write

1.000 token input ditambah 1.000 token output berbiaya sekitar ₩4,44.