AnyModelBot logo

AnyModelBot

Open-source AI assistant for WeChat and Feishu/Lark with 70+

Lets people run AI bots on their own computer or server so chats, memory, files and keys stay local, reaching GPT, Claude and Gemini via one key and paying only for usage.

Built by magnus-2 · Freemium

Visit website

AnyModelBot is a Community listing on Launch Llama with 0 total upvotes, founded by magnus-2 who joined in Nov 2026, compared against 6 alternatives.

Community listing

Listing status

0upvotes

Community score

magnus-2 (Joined Nov 2026)

Founder

Launched on Launch Llama Nov 2, 2026 · #75 of 81 that day

AnyModelBot

About

AnyModelBot is an open-source AI assistant with Modelflare's multi-model service built in. With a single key, it gives you access to more than 70 leading models, including Claude, GPT, Grok, Gemini and DeepSeek, right inside the chat apps you already use every day: WeChat and Feishu/Lark.

Most AI assistants make you open yet another app or browser tab. AnyModelBot works the other way around. You add it to WeChat or Feishu/Lark and talk to it like any other contact or teammate. Ask a question, draft a message, summarize a document, plan your day, or hand it a task, and pick whichever model fits the job best. You can switch between Claude, GPT, Grok, Gemini, DeepSeek and dozens of other models without juggling separate accounts, subscriptions or API keys.

AnyModelBot runs locally on your own computer, so you stay in control. Before it takes any sensitive action, it asks for your approval first, which keeps you in charge of what it does on your behalf. The interface supports both English and Chinese, and the service works on any network worldwide, so teams in different countries can use the same setup.

Key Features

  • Runs on desktop, Docker, WeChat, or Feishu with keys kept on your device
  • Delivers operates your computer by opening apps, reading and moving files, and running commands
  • Stores long-term memory as Markdown files you can open and edit on disk
  • Connects GPT, Claude, Gemini and other models through a single Modelflare Key
  • Provides six characters with separate personas, models, and workspaces

Use Cases

  • Moving downloaded invoices into a reimbursement folder after confirming the file moves
  • Turning a day's chat discussion into a daily report saved to memory
  • Running tests or fixing a code bug from WeChat while away from the computer

Who it is for

  • Users who want an AI assistant on WeChat or Feishu without installing a new app
  • Developers who want a local bot to run scripts, tests, and automation
  • People who want AI to organize files and write daily reports on their own computer

Pricing

The app is free and open source, and model usage is billed pay-as-you-go with a public price table. The page shows example monthly usage estimates from about $0.28 to $11.34 per bot.

View AnyModelBot pricing →

FAQ

和 Grok Bot、OpenAI dots、Claude Duet 有什么不同?

它们是同一类常驻的个人 AI 智能体,在后台替你干活,需要你拍板时再来问你。不同的是,AnyModelBot 开源(MIT),跑在你自己的电脑或服务器上;GPT、Claude、Gemini、Grok 等模型都能用,App 免费,模型按用量付费,不绑定某一家的订阅。GitHub 上的 README 有详细对比表。Grok Bot、dots、Claude Duet 分别是 xAI、OpenAI、Anthropic 的产品,AnyModelBot 与它们没有关联。

我的数据会上传到你们的服务器吗?

对话、记忆、文件和密钥保存在你的设备上。模型请求会经 Modelflare 发往你选择的模型厂商;Modelflare 的日志保留期以其文档《Data retention》为准。电脑操作时的屏幕内容会作为输入发给所选模型。

能在手机上用吗?

暂无独立 App;用微信或飞书和它聊。

我没有 Modelflare 账号?

首次启动点「登录 / 注册 Modelflare」,在浏览器里注册一个就行,回到 App 会自动连好,不用复制 Key。Docker 或远程部署改为粘贴 API Key。在中国大陆可在「高级」里选「中国大陆优化」线路。

电脑操作安全吗?

受限 24 个动作 + 白名单;默认自动执行,可改为“改变界面的操作先询问”;屏幕内容会发给模型;建议先在专用机器或账号上试用。

怎么校验下载的文件?

每个下载旁都有 SHA-256;Windows 用 Get-FileHash,macOS / Linux 用 shasum -a 256。GitHub 发布页另附 SHA256SUMS 文件,可以用 sha256sum -c 一次核对。

关掉窗口后机器人还在吗?

关窗口会驻留托盘(可选开机自启);电脑关机或休眠时不在线;要 24 小时在线请用 Docker。

你们收集遥测数据吗?

我们能确认的是:电脑操作驱动(Cua)的遥测已关闭,magpie 的自动更新已关闭。模型请求如何流转见“数据去向图”。代码是 MIT 开源的,欢迎自行审阅。

Alternatives

WhiziFor anyone paying for more than one AI app: GPT, Claude, Gemini and 280+ models in one workspace, answers side by side, one bill from $15.99/mo.ChatbotX OmnichannelFor agencies: ChatbotX is a self-hosted, white-label omnichannel chatbot platform with visual workflows, AI agents, and multi-channel support across WhatsApp, Facebook, Instagram, and more.NavVaultFor AI power users: Search, organize, and export conversations across ChatGPT, Claude, Gemini, and Grok from one sidebar.APIMartFor developers: APIMart is a unified AI API platform providing access to 500+ models including GPT-5 and Claude 4.5 with one API key, saving up to 70% on costs.9xChatFor builders: 9xChat unifies GPT, Claude, Gemini, image generation, and code execution in one workspace with persistent memory so you stop re-explaining projects across tools.OfoxAIFor developers: OfoxAI gives you one API key to access 100+ LLMs through a single OpenAI-compatible endpoint so you stop juggling multiple provider accounts.

New on Launch Llama

Launched this week · This month's launches · All products A–Z

Tags

#self-hosted ai#wechat bot#feishu integration#multi-model access#local memory#pay-per-use ai#desktop automation#privacy-first

Featured in

Ask AI about AnyModelBot

ChatGPT Claude Perplexity Grok

For bots & AI agents

Machine-readable versions of AnyModelBot and Launch Llama. Fetch these instead of parsing this page.

This listing (.md) llms.txt llms-full.txt ai.txt Full catalog (.md) Endpoint index API docs MCP server MCP manifest Developer docs