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
Providers
Beberapa tempat menghosting model yang sama. Permintaan disebarkan berdasarkan prioritas dan kesehatan, dan local GPU selalu didahulukan jika tersedia.
| Provider | Context | Quantization | Max output | Priority | State |
|---|---|---|---|---|---|
| OpenRouter | 131,07 rb | — | — | 50 | unknown |