Ollama v0.19
Massive local model speedup on Apple Silicon with MLX
Visit website →About
Ollama v0.19 rebuilds inference on Apple Silicon on top of MLX, Apple's machine-learning framework, giving significantly faster local performance for coding and agent workloads on Mac.
It also adds NVFP4 support and smarter handling of the context cache, snapshots, reuse, and eviction, so sessions stay responsive as they grow.
It is aimed at developers running open models locally with Ollama who work with long contexts or agent loops and want the speed to be usable for real work, particularly Mac users who will see the largest gains from the MLX rebuild.
AI-assisted draft, human-reviewed before publishing — see how we choose & review tools.
Why we picked it
This Ollama release earns its spot for the context-cache handling specifically — snapshots, reuse, and eviction keeping a long session responsive is worth naming for anyone who's watched a local model slow to a crawl as a coding session's context grew.
Related tools
- 1 Tiden Verifies coding agents built what you asked for — specs as code Trial
- 2 GitGlow Local Git workbench with smart review for your changes and agent edits Freemium
- 3 GitHub Copilot AI coding assistant inside your editor, with a free tier Freemium
- 4 Together Link Run frontier open models in the coding harness you already use Paid
- 5 Octri Turns OpenAPI specs into docs, SDKs in 10 languages, and MCP server Freemium
- 6 FastRouter.ai One OpenAI-compatible API to route across 200+ LLMs with smart failover Freemium