oMLX 0.7.4 adds prefill accelerator using a second Mac for large models
The oMLX project released version 0.7.4.dev1, a development build that lets a second Mac with a faster GPU accelerate the prefill stage for large models that only fit on a high-memory Mac. The project reports prefill throughput improvements of 43-48% for DeepSeek V4.1 on an M5 Ultra and 16% faster batched decode for Qwen3.8-27B on an M5 Max. New TTS models Ming-3.0-tts and Ming-3.0 Tiny show…
Key points
- Prefill accelerator pairs a fast-GPU Mac with a high-memory Mac over Thunderbolt or 5GbE
- DeepSeek V4.1 prefill 43-48% faster on M5 Ultra; Qwen3.8-27B batched decode +16% on M5 Max
- New TTS models Ming-3.0-tts and Tiny cut real-time factor by 52% and 68%
Model pages: GLM 5.3 → · LFM3 →
oMLX 0.7.4 can pair a fast mac as prefill accelerator
github.com · 4 November 2026Loading the full article…
This text was published by github.com. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
1source- Reddit discussionreddit.com
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Generative AI & Models
All →- Tencent releases Hy4 open-weight model, claims top-tier Chinese AI · 2 src
- Seven open specialist models under 10B parameters outperform larger systems in niche tasks · 1 src
- OpenAI’s 2022: the year ChatGPT and AI art sparked debates over usage and ethics · 1 src
- Axion1-0/Axion-1.0-35B · Hugging Face · 1 src
- OpenAI leaked GPT-6.2 Astra, Anthropic leaked Claude Mythos 5.5 · 3 src
Comments
via GitHub Discussions