deepchat
🐬DeepChat - A smart assistant that connects powerful AI to your personal world
What is deepchat?
🐬DeepChat - A smart assistant that connects powerful AI to your personal world
Overview of deepchat
What it is
DeepChat is an open-source, local-first AI agent desktop client. It brings together models, tools, Skills, agent runtimes, Tape, and long-running sessions in one desktop app. It supports cloud APIs like OpenAI, Gemini, Anthropic, and locally deployed Ollama models. Sessions follow the Tape.systems philosophy to keep processes recoverable, traceable, and inspectable. It also provides MCP support, installable Skills, ACP agent integration, and remote control for messaging platforms.
Key features
- Local-First Agent Desktop Client: Run DeepChat agents, ACP agents, and remote-ready bots in one local app.
- Tape & Trace: Session Tape records structured work history for recovery and resume; Trace previews show request sequences and token budgets.
- Skills: Install, import, export, and enable reusable Skills per conversation for tasks like code review, documents, frontend work, Office/PDF tasks.
- ACP (Agent Client Protocol) Integration: Run ACP-compatible coding and task agents as selectable models.
- Remote Control: Control DeepChat sessions from Telegram, Feishu/Lark, QQBot, Discord, and WeChat iLink.
- Strong MCP Support: Supports Resources, Prompts, Tools, multiple transports, inMemory services, and one-click installation.
- Multi-Platform Support: Windows, macOS, Linux.
Use cases
- Code review and development tasks using installable Skills.
- Document processing and Office/PDF tasks.
- Running ACP-compatible coding agents for structured plans and tool calls.
- Remote control of AI sessions from messaging apps like Telegram, Discord, and WeChat.
- Using multiple cloud LLMs (DeepSeek, OpenAI, Gemini, etc.) and local Ollama models in one application.
- Privacy-focused local data storage with network proxy support.
Getting started
Install via Homebrew on macOS:
brew install --cask deepchat
Alternatively, download the latest release from GitHub Releases for Windows (.exe), macOS (.dmg), or Linux (.AppImage or .deb). After installation, launch the app, click the settings icon, select "Model Providers", add API keys or configure local Ollama, then create a new conversation and select a model.
License and pricing
DeepChat is open source under the Apache License 2.0. No pricing information is provided.
Based on the project's public README and repository metadata.