Twinny

twinny is a self hosted AI coding assistant for VS Code that works with models the user runs. It provides autocomplete, chat, inline edit and code review, and connects to Ollama, llama.cpp, LM Studio or any OpenAI compatible server, whether on a laptop, in a data center or pooled from a team's own computers. The extension is open source under the MIT license, requires no account, sends no telemetry and can be used in an air gapped network. It is aimed at individual developers and teams who want code assistance without code leaving their network. Autocomplete streams fill in the middle suggestions as ghost text and is tuned for small models. Chat can attach a file, symbol, the problems panel, the git diff, the terminal or a search of the workspace index with the @ key. Inline edit, opened with Ctrl+I, shows the change as a diff that can be accepted or rejected per hunk. A hybrid keyword and semantic workspace index supports the chat. For teams, one gateway serves everyone, with a key per developer and recording off by default. Pricing is published. The free tier costs $0 and covers individuals and teams of up to five seats on a gateway, with community support. Team seats are $6 per seat per month billed yearly, which is $72 per year for each seat beyond the free five, and include a 30 day trial with no card required. The enterprise tier is $10 per seat per month billed yearly, from 50 seats. Payment is by card, and a license token is issued for the gateway.