Running local models is good now

	Running local models is good now(vickiboykis.com)
	1556 points by jfb 46 days ago \| 597 comments
	tl;dr: Local LLMs have crossed a usability threshold, with GPT-OSS and Gemma 3 (referred to as "Gemma 4" in the post) enabling agentic coding workflows on a 64GB M2 Mac at roughly 75% of frontier model performance. The author runs Pi as an agent harness against LM Studio in a Docker sandbox, using it for refactoring, unit tests, and bootstrapping repos—though slow inference and small context windows still rule out production use.
	HN Discussion: ↓Local models are still painful to run with quantization and memory tradeoffs lobotomizing them ↑Local models are genuinely preferable to frontier models for actual daily use ↑Rise of local models threatens the pricing power of frontier model providers ↓Local models are far behind frontier models and the article oversells their capabilities ~Local models work well when paired with frontier planning or used for narrow, well-defined tasks