Context
262K
Input
$0.10/M
Output
$0.30/M
Aliases:qwen3-30b-a3b-instruct-2507qwen3-30b-a3b-2507
Alibaba's Qwen3 30B mixture-of-experts model (3B active, 128 experts) with a native 262K context. Operates in non-thinking mode only and emits no reasoning blocks.
nebius/qwen3-30b-a3b-instructQuantfp8
- Input
- $0.10/M
- Output
- $0.30/M
- Context
- 262K
- Max out
- 33K
- Caching
- Not offered
- Thinking out
- $0.30/M