Verified project record

ThinkInAIXYZ/deepchat

DeepChat is a local-first desktop application that unifies cloud and local AI models with agent skills, MCP support, and messaging-based remote control. It is built for users who need a privacy-focused client to connect multiple models and external agent runtimes to personal workflows.

Project overview

Project type: MCP · AI Agent · Model Runtime
Use cases: Documents & Office · Knowledge Q&A · Coding & Development
Deployment: Refer to project documentation
License: Apache-2.0

Best for

Developers and general users seeking a local-first, desktop GUI application to unify various cloud and local AI models.

Key capabilities

Session Tape records structured work history for recovery and resume, while Trace previews expose request sequences, metadata, and token budgets.
Allows installing and enabling reusable Skills per conversation for specific tasks. Supports importing/exporting with various compatible tools.
Native support for Agent Client Protocol to integrate external agent runtimes as first-class entries in the model selector.
Allows controlling DeepChat sessions from Telegram, Feishu/Lark, QQBot, Discord, and WeChat iLink, including binding endpoints, switching models, and answering prompts.
Full support for MCP Resources, Prompts, and Tools across multiple transports like StreamableHTTP, SSE, and Stdio. Includes a built-in Node.js runtime and inMemory services.
Supports mainstream cloud LLMs (OpenAI, Gemini, Claude, DeepSeek, etc.) and local Ollama models within a single application.
Built-in integrations with leading search APIs and support for simulating user web browsing to read search engines like a human.
Supports multi-window/tab architecture, complete Markdown rendering, Artifacts rendering, message retries, and conversation forking.

Limitations and risks

Homebrew installation is restricted to macOS users.
GPU and minimum hardware requirements are not documented. Telemetry practices are not documented.

Getting started

Download a pre-built installer from GitHub Releases or Homebrew. No code compilation is required. After installation, launch the application and configure model providers or a local Ollama instance.

Evidence and sources

README: DeepChat is an open-source, local-first AI agent desktop client with rich agent capabilities, designed around the Tape.systems philosophy, with support for MCP, Skills, ACP, and r…
README: DeepChat is a powerful open-source, local-first AI agent desktop client that brings together models, tools, Skills, agent runtimes, Tape, and long-running sessions in one desktop…
README: Compared to other AI tools, DeepChat offers the following unique advantages: - Local-First Agent Desktop Client: Run DeepChat agents, ACP agents, and remote-ready bots in one loca…
README: - 🤖 Local-First Agent Desktop Client - Select DeepChat, ACP, and remote-capable agents from one model-like entry point - Run long-lived sessions with project folders, permission m…
README: DeepChat's session Tape follows the Tape.systems philosophy and keeps agent work recoverable and inspectable. Trace previews expose request sequences, provider/model metadata, Tap…

Reviewed at: 2026-07-16T13:41:54Z