Observed arrival · 2026-09-10
Just Fucking Use Open Code
A profane, practical guide to running OpenCode with local Ollama models instead of sending code to hosted AI services.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
Field notes
The guide’s workflow begins with Ollama, downloads Qwen 2.5 Coder models, and points OpenCode at a local OpenAI-compatible endpoint on port 11434. It distinguishes a smaller 1.5B model from a 7B model that requires more RAM, and separately recommends nomic-embed-text for codebase search. The page also describes Plan and Build modes, /undo, and an /init command that writes AGENTS.md for repository context.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue