Unverified50% confidenceFactExact time
Anifer supports MTP (Multi-Token Prediction) speculative decoding with an acceptance rate between 46% and 49%, and also supports D Flash speculative decoding but with higher VRAM usage.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/16/2026
First Seen
Valid until: 12/15/2026
Sources
Related Entities
Related Claims
UnverifiedIn DeepSeek V4 Flash deployments, DiSpark improved single-user generation speed by 60% to 85%, and the Pro series by 57% to 78%, while maintaining equivalent inference quality63% similarUnverifiedv4.1 Flash introduces a mechanism called Ngram, where 2 out of 40 layers contain conditional memory tables, with 196 billion parameters used for a lookup table that matches sequences of up to 4 tokens against a vocabulary of 16 million.58% similarUnverifiedFlashAttention通过优化GPU内存访问模式提升计算效率，ALiBi、RoPE等位置编码改进赋予模型更强的长程外推能力57% similarUnverifiedFlashInfer MNNVL pure AllReduce achieves up to +6.9% performance improvement for DeepSeek-V4-Flash TP4 decoding on Blackwell architecture at small batch sizes.56% similarVerifiedThe Flash version is positioned as a lighter, faster inference option suited for latency-sensitive or resource-constrained scenarios, while the PRO version targets a higher capability ceiling with greater demands on VRAM and compute56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/924351API
curl https://kongchang.com/api/v1/knowledge/claims/924351MCP
get_claim(id=924351)