On Hugging Face · first seen 14 October 2026, 15:11 UTC
Ling-3.1-flash: 560B-parameter MoE model with 1M context
A huge open MoE with 1M context pushes long-context open models further.
An open Hugging Face model with 560B total parameters, 25B active per token and a 1M-token context window. It follows Ling 3.0, which was 127B.
inclusionAIOpenOpen weights
- Name
- Ling-3.1-flash
- Params
- 560B total, 25B active
- Context
- 1M