Unverified60% confidenceFactExact time
The DeepSeek team fine-tuned on DeepSeek v4 flash and developed their own visual encoder, DeepSeek ViT.
2
Sources
60%
Confidence
Medium-term (~90 days)
Relevance
7/24/2026
First Seen
Valid until: 10/22/2026
Sources
DeepSeek神秘撤稿论文：用视觉基元突破AI多模态推理瓶颈
bilibili枫叶边城7/11/2026
Related Claims
VerifiedDeepSeek has released an experimental multimodal model, DeepSeek V4 Flash Vision EXP, which performs excellently in both text and multimodal tasks.80% similarUnverifiedDeepSeek kernel engineers wrote the core attention kernel for DeepSeek V4.1 Flash.78% similarUnverifiedDeepSeek released DeepSeek-V4.1-Flash, a method that significantly compresses the memory required for KV cache.78% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/608999API
curl https://kongchang.com/api/v1/knowledge/claims/608999MCP
get_claim(id=608999)