Unverified50% confidenceBenchmarkExact time
FlashInfer MNNVL pure AllReduce achieves up to +6.9% performance improvement for DeepSeek-V4-Flash TP4 decoding on Blackwell architecture at small batch sizes.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/5/2026
First Seen
Valid until: 12/4/2026
Sources
Related Entities
Related Claims
UnverifiedThe FlashInfer pure allreduce feature is automatically enabled for DeepSeek-V3/V3.2/V4, and other models can enable it manually via --enable-flashinfer-pure-allreduce.76% similarUnverifiedDeepSeek V4 Flash is a distilled or streamlined version of Pro that significantly reduces inference overhead while retaining core programming capabilities.72% similarVerifiedDeepSeek has launched the experimental multimodal model V4-Flash-Vision-EXP on its API platform, with text capabilities on par with V4 Flash and multimodal agent benchmark performance close to Opus 4.8 levels.70% similarUnverifiedDeepSeek-V4-Flash-Vision-Exp releases weights under the MIT license and provides reference implementations for the Tensor and Tokenizer as well as minimalist PyTorch inference code.69% similarUnverifiedDeepSeek在7月的更新日志中将一个内部Harness（称为「极简模式」）归功为V4 Flash编程基准评分的幕后功臣68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/860551API
curl https://kongchang.com/api/v1/knowledge/claims/860551MCP
get_claim(id=860551)