APITry the API →
HermesHermes Agent Docs
API
Back to News
modelLlmsGpus

Atomic Chat Runs a 177B Qwen3.8-Flash-Next on a 64 GB Mac

A 177B parameter Qwen4-preview MoE now runs on a 64GB MacBook at 36 tokens per second, thanks to an SSD-paged n-gram table.

Atomic Chat Runs a 177B Qwen3.8-Flash-Next on a 64 GB Mac
Source
AlphaSignal
Published
Author
AlphaSignal Newsroom
Read
1 min read

A 177B parameter Qwen4-preview MoE now runs on a 64GB MacBook at 36 tokens per second, thanks to an SSD-paged n-gram table.

Reporting is indexed from AlphaSignal. Rights remain with the original publisher and cited sources.

Read original report

Next reads

SC117 Strips Qwen3.8-Flash-Next Refusals Without Touching 1,079 Tensors
AlphaSignal · model

SC117 Strips Qwen3.8-Flash-Next Refusals Without Touching 1,079 Tensors

AutoTrust AI Turns Gemma 4 Into a 45ms Calibrated Decision API
AlphaSignal · model

AutoTrust AI Turns Gemma 4 Into a 45ms Calibrated Decision API

Liquid AI's d1-3B Makes Structured AI Decisions in 8 Milliseconds Without Generating Text
Liquid AI · model

Liquid AI's d1-3B Makes Structured AI Decisions in 8 Milliseconds Without Generating Text