Verified85% confidenceFactExact time
Qwen3.8-Flash-Next adopts a Mixture of Experts (MoE) architecture with 125B total parameters and 6B activated parameters.
5
Sources
85%
Confidence
Medium-term (~90 days)
Relevance
9/5/2026
First Seen
Valid until: 12/4/2026
Sources
Related Entities
Related Claims
UnverifiedAlibaba's Tongyi Qianwen officially released the Qwen3.8-Flash-Next (Qianwen 308 Flash NEX) model, with a total parameter count of 181.5B.78% similarVerifiedQwen3 Max adopts a MoE (Mixture of Experts) architecture.78% similarVerifiedQwen 3.8 Flash Next adopts a Mixture of Experts (MoE) architecture.77% similarUnverifiedQwen3.8-Flash-Next adopts a hybrid attention mechanism of GDN plus QSA, with a total of 125B parameters plus 51B N-gram embeddings, activating only approximately 6B parameters per token.76% similarUnverifiedOn a single RX 7900 XTX 24G GPU, the Strata engine can fully run the approximately 125B-parameter Qwen3.8 Flash Next MoE model, supporting both text inference and image recognition.76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/858929API
curl https://kongchang.com/api/v1/knowledge/claims/858929MCP
get_claim(id=858929)