Thinking Machines' Inkling-Small Squeezes a 276B Multimodal Model onto One GPU
Thinking Machines Lab's 276B-parameter MoE matches its 975B flagship across reasoning, coding, and multimodal tasks while using 12B active parameters per token.
Source
AlphaSignal
Published
Author
AlphaSignal Newsroom
Read
1 min read
Thinking Machines Lab's 276B-parameter MoE matches its 975B flagship across reasoning, coding, and multimodal tasks while using 12B active parameters per token.
Reporting is indexed from AlphaSignal. Rights remain with the original publisher and cited sources.
