Sandbox
@succlz123/DeepCo

Compose Multiplatform chat client for LLM providers

Deep-Co is a chat client for LLMs built with Compose Multiplatform. It connects to providers like OpenRouter, Anthropic, Grok, OpenAI, DeepSeek, Coze, Dify, Gemini, and OpenAI-compatible APIs, and it can also use local models through LM Studio or Ollama. The app includes chat history, prompt management, SillyTavern character support, MCP server setup, and Edge TTS. The repository is organized as a multi-platform app with desktop, Android, iOS, and shared code modules.

38 stars7 forksKotlinUpdated 10mo ago
Who it's for

People who want one chat app for several LLM providers and local models.

What it delivers

You can chat with different models, manage prompts, and connect MCP servers in one app instead of switching tools.

What it does

Multi-provider chat

Connects to OpenRouter, Anthropic, Grok, OpenAI, DeepSeek, Coze, Dify, and Google Gemini, plus any OpenAI-compatible API.

Local model support

Can use native models through LM Studio or Ollama.

Chat history

Supports streamed and complete chat responses with saved history.

Prompt management

Lets you create and manage your own prompts.

SillyTavern character adaptation

Supports PNG and JSON character files from SillyTavern.

MCP support

Lets you configure MCP servers and use model context protocol connections.

TTS

Includes Edge API text-to-speech support.

Cross-platform UI

Built with Compose Multiplatform for desktop and mobile targets.

How to get it

  1. 1Run
    brew install uv
    brew install node
  2. 2Run
    winget install --id=astral-sh.uv  -e
    winget install OpenJS.NodeJS.LTS
  3. 3Run
    ./gradlew :desktopApp:run
  4. 4Run
    ./gradlew :desktop:packageDistributionForCurrentOS
    # outputs are written to desktopApp/build/compose/binaries
  5. 5Run
    ./gradlew :androidApp:installDebug
  6. 6Run
    ./gradlew clean :androidApp:assembleRelease
    # outputs are written to androidApp/build/outputs/apk/release

README

Deep-Co

icon

windows macos linux
android iOS
kotlin compose
stars gpl release

A Chat Client for LLMs, written in Compose Multiplatform. Target supports API providers such as OpenRouter, Anthropic, Grok, OpenAI, DeepSeek, Coze, Dify, Google Gemini, etc. You can also configure any OpenAI-compatible API or use native models via LM Studio/Ollama.

Release

v1.0.6

Feature

  • Desktop Platform Support(Windows/MacOS/Linux)
  • Mobile Platform Support(Android/iOS)
  • Chat(Stream&Complete) / Chat History
  • Chat Messages Export / Chat Translate Server
  • Prompt Management / User Define
  • SillyTavern Character Adaptation(PNG&JSON)
  • DeepSeek LLM / Grok LLM / Google Gemini LLM
  • Claude LLM / OpenAI LLM / OLLama LLM
  • Online API polling
  • MCP Support
  • MCP Server Market
  • RAG
  • TTS(Edge API)
  • i18n(Chinese/English) / App Color Theme / App Dark&Light Theme

Chat With LLMs

1

Config Your LLMs API Key

2

Prompt Management

4

Chat With Tavern Character

6

User Management

5

Config MCP Servers

3

Setting

7

Model Context Protocol (MCP) ENV

MacOS

brew install uv
brew install node

windows

winget install --id=astral-sh.uv  -e
winget install OpenJS.NodeJS.LTS

Build

Run desktop via Gradle

./gradlew :desktopApp:run

Building desktop distribution

./gradlew :desktop:packageDistributionForCurrentOS
# outputs are written to desktopApp/build/compose/binaries

Run Android via Gradle

./gradlew :androidApp:installDebug

Building Android distribution

./gradlew clean :androidApp:assembleRelease
# outputs are written to androidApp/build/outputs/apk/release

Thanks

Files in the repo

Repository payload16 top-level entries
  • .github
  • androidApp
  • desktopApp
  • gradle
  • iosApp
  • screenshots
  • shared
  • .gitignore
  • .gitmodules
  • build.gradle.kts
  • compose-screen
  • gradle.properties
  • gradlew
  • gradlew.bat
  • README.md
  • settings.gradle.kts

Discussion (0)

Ask about usage, or say what you built with it

Sign in to join the discussion.

No comments yet. Be the first to say what this is good for.

More other

HKUDS/
Vibe-Trading

"Vibe-Trading: Your Personal Trading Agent"

33k
tinyhumansai/
openhuman

OpenHuman is an open source personal AI for Mac, Windows and Linux — local-first memory, agent orchestration, and deep research.

40k

Your Personal AI Assistant; easy to install, deploy on your own machine or on the cloud; supports multiple chat apps with easily extensible capabilities.

35k

Open-source Claude Design alternative. One-click import your Claude Code / Codex API key. Prompt → prototype / slides / PDF. Multi-model (Claude, GPT, Gemini, Kimi, GLM, Ollama). BYOK, local-first, MIT.

7.9k

🤯 LobeHub is your Chief Agent Operator, organizing your agents into 7×24 operations by hiring, scheduling, and reporting on your entire AI team.

82k