Community poll of 584 ballots crowns oMLX (57%) the top MLX engine on Apple Silicon, ahead of MLX-Serve and LM Studio
andrejusb · x · 2026-09-26
ivanfioravanti ran a community poll on which MLX engines people actually use on Apple Silicon, drawing 584 ballots.
- Results: oMLX leads decisively at 57%, followed by MLX-Serve (38.4%), LM Studio (33.9%), MTPLX (28.6%), and Ollama (15.1%), across 22 total engines.
- By chip: 87% of M3 Ultra owners pick oMLX; base M4 users favor Ollama and Unsloth; base M5 users lean heavily on LM Studio (82%).
- Sample: 370 respondents shared their chip, dominated by M5 Max (26.8%) and M4 Pro (12.2%).
The data suggests oMLX has become the de facto favorite among high-end Mac local-inference users, while entry-level device users show notably different tool preferences.
More from Infra
- Niantic opens Places Library: 100 real-world 3D scenes for embodied AI training — Scobleizer · 2026-09-26
- umbrelOS 2.0 Launches: Turn a Spare Computer Into a Private Cloud via USB — JosephJacks_ · 2026-09-26
- Transformers.js monthly downloads grew from 1.7M to 10M in a year — nicodotdev · 2026-09-26
- 85.6 tok/s on a single RTX 5090: local Qwen3 27B with MTP speedup — comperr · 2026-09-26
- Chart claims $6.5 trillion in current and future liabilities between LLM labs and hyperscalers — AIFlow_ML · 2026-09-26
- 128GB local LLM server forges 80M tokens a month on just $6 of electricity — No-Fuel-9202 · 2026-09-26