Verified65% confidenceFactExact time
DeepSeek-V4.1-Flash's KV cache is compressed to 890 bytes per token, only one-quarter that of V4 Flash.
3
Sources
65%
Confidence
Long-term
Relevance
9/16/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedFrom V1 to V4.1 Flash, the single-token cache was compressed from 389,000 bytes to 890 bytes, a total compression of 437 times.78% similarUnverifiedAfter compression, the DeepSeek V4 Flash model size was reduced from nearly 500 GB to 80.8 GB.74% similarUnverifiedThe UD-Q8_K_XL quantized version of DeepSeek V4 Flash has a file size of 162 GB, uses MXFP4+BF16 mixed precision, and achieves completely lossless quality.73% similarVerifiedV4 Flash's cache size is only 7% of that of V3.2.73% similarUnverifiedDeepSeek V4 Flash is a cloud-based model focused on high throughput, with an inference speed of approximately 140 tokens per second.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/924480API
curl https://kongchang.com/api/v1/knowledge/claims/924480MCP
get_claim(id=924480)