{"name":"omlx","repo":"jundot/omlx","url":"https://midorreal.com/project/jundot-omlx","repository":"https://github.com/jundot/omlx","description":"LLM inference server with continuous batching & SSD caching for Apple Silicon — managed from the macOS menu bar","facts":{"license":{"value":"Apache-2.0","source":"https://github.com/jundot/omlx","checked":"22 Nov 2026"},"language":{"value":"Python","source":"https://github.com/jundot/omlx","checked":"22 Nov 2026"},"homepage":{"value":"https://omlx.ai","source":"https://omlx.ai","checked":"22 Nov 2026"},"last_release":{"value":"v0.8.0, 18 Nov 2026","source":"https://github.com/jundot/omlx/releases","checked":"22 Nov 2026"},"releases":{"value":"139","source":"https://github.com/jundot/omlx/releases","checked":"20 Nov 2026"},"contributors":{"value":"375","source":"https://github.com/jundot/omlx/graphs/contributors","checked":"20 Nov 2026"},"archived":{"value":"no","source":"https://github.com/jundot/omlx","checked":"22 Nov 2026"}},"stars":23820.0,"writeup":{"what_it_is":"oMLX is an LLM inference server for Apple Silicon Macs. It uses continuous batching and a tiered KV cache (RAM hot tier and SSD cold tier), and is managed from the macOS menu bar. It serves text LLMs, vision-language models, embeddings and rerankers through an OpenAI-compatible API.","audience":"Mac users running local LLMs for everyday use and coding tools","claims":[{"kind":"specific","claim":"Runs on Apple Silicon Macs","excerpt":"platform-Apple%20Silicon-black","status":"not_checked"},{"kind":"specific","claim":"Requires macOS 15.0+ and Python 3.11–3.13","excerpt":"Requires macOS 15.0+ (Sequoia), Python 3.11–3.13","status":"not_checked"},{"kind":"specific","claim":"Persists KV cache to SSD so cached prefixes can be restored after a server restart","excerpt":"even after a server restart","status":"not_checked"},{"kind":"specific","claim":"Handles concurrent requests via continuous batching","excerpt":"Handles concurrent requests through mlx-lm's BatchGenerator","status":"not_checked"},{"kind":"broad","claim":"Makes local LLMs practical for real coding work","excerpt":"making local LLMs practical for real coding work","status":"too_broad"}],"alternatives":[],"written_by":"AI, from the project's README","date":"2026-11-14"},"why_now":null,"owner_supplied":null,"score":{"verdict":"real","hype":92,"reality":98,"gap":-5,"momentum":0.9,"confidence":60,"version":"score-2.0","rule":"verdict-2.0","calculated_at":"2026-11-22T08:00:00+00:00","verdict_source":"rule"},"early_read":null,"changes":[{"date":"2026-11-18T06:40:09+00:00","text":"Latest release v0.8.0","source":"https://github.com/jundot/omlx/releases"},{"date":"2026-02-13T14:50:53+00:00","text":"First release","source":"https://github.com/jundot/omlx/releases"},{"date":"2026-02-13T14:13:27+00:00","text":"Repository created","source":"https://github.com/jundot/omlx"}],"cite":"omlx is a GitHub project at https://github.com/jundot/omlx written mainly in Python, described there as \"LLM inference server with continuous batching & SSD caching for Apple Silicon - managed from the macOS menu bar\".","attribution":"Data from Mid or Real, https://midorreal.com/project/jundot-omlx, checked 22 November 2026. Attribution with a link is required.","still_real":null}