Unverified50% confidenceReleaseExact time
Meta simultaneously released a draft model called D-Flash, used for speculative decoding to accelerate generation.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/16/2026
First Seen
Valid until: 11/14/2026
Sources
实测Meta开源Muse Glimmer：偏科的Agent专用模型
bilibili碳基生物退役指南8/11/2026
Related Claims
UnverifiedMeta has open-sourced PyTorch and developed its own MTIA inference chip.66% similarUnverifiedMeta开源了名为Muse Glimmer的30B规模多模态Agent模型64% similarUnverifiedWithin the first hour after the model's release, llama.cpp merged the adaptation PR into the main branch, including support for the model, visual input, and D-Flash.63% similarUnverifiedD-My建立在扩散模型（Diffusion Model）与Transformer架构的融合基础上61% similarUnverifiedLlama.cpp was originally created with the goal of running Meta's LLaMA model on pure CPU environments60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/755344API
curl https://kongchang.com/api/v1/knowledge/claims/755344MCP
get_claim(id=755344)