
by TabbyML
Tabby is an open-source, self-hosted AI coding assistant offering code completion, an answer engine and inline chat. It runs on your own infrastructure without needing an external database or cloud service.
PluginFreemium
Step 1: Start a Tabby server, for example with the Docker command from the README, choosing a completion model and a chat model.
Step 2: Open the Tabby web interface and complete the initial setup.
Step 3: Install the Tabby extension for your editor, such as VS Code or a JetBrains IDE.
Step 4: Point the extension at your Tabby server.
Step 5: Use completions as you type, ask questions through the Answer Engine, and use inline chat for edits.
Step 6: Add context providers so answers draw on your documentation and repositories.
No. The README describes Tabby as self-contained, with no need for a DBMS or cloud service.
The site lists VS Code, Neovim and IntelliJ-family IDEs including PyCharm, WebStorm, GoLand and Android Studio, among others.
Yes. The README says Tabby supports consumer-grade GPUs, and its Docker example starts a server with the GPU enabled.
Tabby is described as compatible with major coding LLMs, and the docs include a models directory of recommended models. You can combine your preferred models.
BetterMemory MCPBetterMemory is an MCP memory server for coding agents. It checks stored facts against the filesystem and git before the model relies on them, so outdated memories are flagged.