Unverified50% confidenceFactExact time
Stratum focuses on adapting to very large MOE models such as Qwen 3.8 Flash, Next and DeepSeek V4.1 Flash.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/31/2026
First Seen
Valid until: 1/29/2027
Sources
Related Entities
Related Claims
VerifiedDeepSeek V4 Flash is an MoE multimodal model with a total parameter count of 284B.76% similarVerifiedThis update adds support for the official model ID deepseek-flash; new sessions default to V4.1 Flash, but explicitly configured models take higher priority.74% similarUnverifiedOn the DeepSWE v1.1 benchmark, 3.8 Flash surpasses most frontier models with larger parameter counts.74% similarVerifiedQwen 3.8 Flash Next adopts a Mixture of Experts (MoE) architecture.73% similarUnverifiedDeepSeek V4.1 Flash is on par with GLM 5.3 and Kimi K3 on several key benchmark tests.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1120200API
curl https://kongchang.com/api/v1/knowledge/claims/1120200MCP
get_claim(id=1120200)