Unverified50% confidenceTradeoffExact time
There is a smaller approximately 9GB version of the Qwen3.8 27B model on Hugging Face, but its quality will be degraded.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/11/2026
First Seen
Valid until: 12/10/2026
Sources
Related Entities
Related Claims
UnverifiedMiniMax M3 的 HuggingFace BF16 版本仓库约占 854GB72% similarUnverifiedQwen 3.6 35B A3B is fully open-source, with weights available on Hugging Face at approximately 72 GB in size.66% similarUnverifiedFor a model of approximately 160 GB (such as the UD-Q3_K_XL quantized version of Qwen3.5-397B), running the lens requires approximately 20 GB of additional memory.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/907317API
curl https://kongchang.com/api/v1/knowledge/claims/907317MCP
get_claim(id=907317)