Sidekick

A native macOS app that allows users to chat with a local LLM that can respond with information from files, folders and websites on your Mac without installing

What is Sidekick?

A native macOS app that allows users to chat with a local LLM that can respond with information from files, folders and websites on your Mac without installing any other software. Powered by llama.cpp.

Overview of Sidekick

What it is

Sidekick is a native macOS application that allows users to chat with a local large language model (LLM) that can respond with information from files, folders and websites on the user's Mac without installing any other software. All conversations happen offline, and data stays secure. Sidekick is a local-first application with a built-in inference engine for local models, while also accommodating OpenAI compatible APIs for additional model options. It is powered by llama.cpp.

Key features

  • Resource Use: Sidekick accesses files, folders, and websites from configurable experts using RAG (Retrieval Augmented Generation). Users can also give Sidekick access to files by dragging them into the input field. Sidekick can respond with the latest information using web search.
  • Bring Your Own API Key: In addition to local models, Sidekick allows users to bring their own key for OpenAI compatible APIs. It ships with built-in presets for providers including OpenAI, Anthropic, Google AI Studio, DeepSeek, Groq, MiniMax, Mistral, xAI, and more.
  • Function Calling: Sidekick can call functions to boost mathematical and logical capabilities and to execute actions. Functions are called sequentially in a loop until a result is obtained.
  • Deep Research: A specific agent implemented in Sidekick to handle long horizon, multi-step research tasks. It reads 50-80 webpages and synthesizes information to prepare a research report.
  • Memory: Sidekick can remember helpful information between conversations, making responses more relevant and personalized.
  • Canvas: Create, edit and preview websites, code and other textual content using Canvas. Users can select parts of the text and prompt the chatbot to perform selective edits.
  • Image Generation: Sidekick can generate images from text using a built-in CoreML model that automatically identifies image generation prompts. Image generation is available on macOS 15.2 or above and requires Apple Intelligence.

Use cases

Sidekick can be used for research, such as collecting evidence for a history paper about interactions between Aztecs and Spanish troops. A user can ask a question and Sidekick responds with direct quotes with page numbers and a brief analysis. It can also be used for mathematical and logical calculations, drafting emails (e.g., finding a birthday and email address from contacts and creating a draft in the default email client), and generating images from text.

Getting started

Requirements: A Mac with Apple Silicon and RAM ≥ 8 GB.Installation via Homebrew:bashbrew install --cask arcadi4/tap/sidekick-aiFor download and setup, follow the guide at the repository.

License and pricing

Sidekick is open source under the MIT license.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/johnbean393sidekick.svg)](https://yunodb.com/tools/johnbean393sidekick?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/johnbean393sidekick?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/johnbean393sidekick.svg" alt="Featured on YunoDB" height="20"></a>