|
26 | 26 |
|
27 | 27 | --- |
28 | 28 |
|
29 | | -## What is this? |
| 29 | +## What is OpenWhale? |
30 | 30 |
|
31 | | -OpenWhale is an AI assistant that can take action. It connects to multiple AI models, talks on your behalf through WhatsApp/Telegram/Discord, browses the web, executes code, manages your calendar, and basically does whatever you need it to do. |
| 31 | +OpenWhale is a **self-extensible AI assistant** that doesn't just chat β it takes action. Connect it to multiple AI models, let it talk on your behalf through WhatsApp/Telegram/Discord/Twitter, browse the web with real browser automation, execute code, manage your calendar, send emails, and basically do whatever you need it to do. |
32 | 32 |
|
33 | | -Think of it as Claude, GPT-4, or DeepSeek with arms and legs. |
| 33 | +Think of it as **Claude, GPT-4, or DeepSeek with arms and legs**. |
| 34 | + |
| 35 | +<p align="center"> |
| 36 | + <img src="images/dashboardhome.png" alt="OpenWhale Dashboard" width="800" /> |
| 37 | +</p> |
| 38 | + |
| 39 | +--- |
| 40 | + |
| 41 | +## β¨ Core Features |
| 42 | + |
| 43 | +### π€ 8 AI Providers |
| 44 | +Switch between any major AI model on the fly. One assistant, unlimited model choices. |
| 45 | + |
| 46 | +| Provider | Top Models | |
| 47 | +|----------|------------| |
| 48 | +| **Anthropic** | Claude Sonnet 5, Opus 4.5, Sonnet 4.5 | |
| 49 | +| **OpenAI** | GPT-5.2, GPT-5, GPT-4o, o4-mini | |
| 50 | +| **Google Gemini** | Gemini 3 Pro, Gemini 2.5 Pro | |
| 51 | +| **Qwen** | Qwen3-Max, QwQ-Plus, Qwen3-Coder | |
| 52 | +| **DeepSeek** | DeepSeek Chat, DeepSeek Coder, DeepSeek Reasoner | |
| 53 | +| **Groq** | Llama 3.3 70B, Mixtral 8x7B | |
| 54 | +| **Together AI** | Any open-source model | |
| 55 | +| **Ollama** | Local models (Llama, Mistral, Phi) β no API key! | |
| 56 | + |
| 57 | +<p align="center"> |
| 58 | + <img src="images/aiproviders.png" alt="AI Providers" width="600" /> |
| 59 | +</p> |
| 60 | + |
| 61 | +--- |
| 62 | + |
| 63 | +### π¬ 4 Messaging Channels |
| 64 | +Your AI responds across all platforms β unified inbox, one brain. |
| 65 | + |
| 66 | +| Channel | How it Works | |
| 67 | +|---------|--------------| |
| 68 | +| **WhatsApp** | Scan QR code, uses your personal account | |
| 69 | +| **Telegram** | Create a bot with @BotFather | |
| 70 | +| **Discord** | Add bot to your server | |
| 71 | +| **Twitter/X** | Auto-reply to mentions via bird CLI (cookie auth, no API!) | |
| 72 | + |
| 73 | +--- |
| 74 | + |
| 75 | +### π οΈ 15 Built-in Tools |
| 76 | +The AI can actually DO things, not just talk about them. |
| 77 | + |
| 78 | +| Tool | Capability | |
| 79 | +|------|------------| |
| 80 | +| **exec** | Run any shell command on your machine | |
| 81 | +| **file** | Read, write, list files and directories | |
| 82 | +| **browser** | Full web automation β navigate, click, type, screenshot | |
| 83 | +| **screenshot** | Capture your screen, AI analyzes what it sees | |
| 84 | +| **code_exec** | Run Python/JavaScript/TypeScript code on-the-fly | |
| 85 | +| **web_fetch** | Fetch any URL, parse APIs and web pages | |
| 86 | +| **memory** | Remember things across conversations forever | |
| 87 | +| **cron** | Schedule tasks to run at specific times | |
| 88 | +| **canvas** | Generate and manipulate images | |
| 89 | +| **tts** | Text-to-speech (AI speaks out loud) | |
| 90 | +| **image** | Analyze and process images with vision | |
| 91 | +| **nodes** | Structured data and knowledge graphs | |
| 92 | +| **camera** | Take photos/video from connected cameras | |
| 93 | +| **location** | Get current geolocation | |
| 94 | +| **extend** | Create self-extensions for automated workflows | |
| 95 | + |
| 96 | +<p align="center"> |
| 97 | + <img src="images/tools.png" alt="Available Tools" width="800" /> |
| 98 | +</p> |
| 99 | + |
| 100 | +--- |
| 101 | + |
| 102 | +### π 9 API Skills (External Integrations) |
| 103 | +Connect to your favorite services β the AI handles the API calls. |
| 104 | + |
| 105 | +| Skill | What it Does | |
| 106 | +|-------|--------------| |
| 107 | +| **GitHub** | Manage repos, issues, PRs, commits | |
| 108 | +| **Notion** | Search, create, update pages and databases | |
| 109 | +| **Google Calendar** | View and create events | |
| 110 | +| **Gmail** | Read, send, and search emails | |
| 111 | +| **Google Drive** | Upload, download, list files | |
| 112 | +| **Google Tasks** | Manage your to-do lists | |
| 113 | +| **Weather** | Current conditions and forecasts | |
| 114 | +| **Spotify** | Control playback, search music, manage playlists | |
| 115 | +| **Trello** | Manage boards, lists, and cards | |
| 116 | +| **1Password** | Securely fetch passwords and secrets | |
| 117 | +| **Apple Notes/Reminders** | Native macOS integration | |
| 118 | +| **Twitter/X** | Post, search, timeline, mentions, follow | |
| 119 | + |
| 120 | +<p align="center"> |
| 121 | + <img src="images/api-skills.png" alt="API Skills" width="800" /> |
| 122 | +</p> |
| 123 | + |
| 124 | +--- |
| 125 | + |
| 126 | +### π 50+ Markdown Skills (Community) |
| 127 | +Community-built skills from [OpenClaw](https://github.com/VoltAgent/awesome-openclaw-skills) β instant expertise in any domain. |
| 128 | + |
| 129 | +- **Frontend**: UI/UX design, React best practices, design systems |
| 130 | +- **Backend**: Architecture patterns, NextJS, Vercel deployment |
| 131 | +- **AI/Video**: ComfyUI, Remotion video toolkit, vision models |
| 132 | +- **Platform**: Apple HIG, Linux service management, Xcode builds |
| 133 | +- **Productivity**: Resume builder, email templates, artifact creation |
| 134 | + |
| 135 | +<p align="center"> |
| 136 | + <img src="images/mdskills.png" alt="Markdown Skills" width="800" /> |
| 137 | +</p> |
| 138 | + |
| 139 | +Create your own skills with the built-in Skill Creator: |
| 140 | + |
| 141 | +<p align="center"> |
| 142 | + <img src="images/skill-creator.png" alt="Skill Creator" width="800" /> |
| 143 | +</p> |
| 144 | + |
| 145 | +--- |
| 146 | + |
| 147 | +### β‘ Self-Extension System |
| 148 | +The AI can create its own automations. Just ask. |
| 149 | + |
| 150 | +- *"Create an extension that checks Bitcoin price every hour and sends it to WhatsApp"* |
| 151 | +- *"Make an extension that reminds me to drink water"* |
| 152 | +- *"Create a daily standup summary extension"* |
| 153 | + |
| 154 | +Extensions support cron scheduling, multi-channel output, and persist across restarts. |
| 155 | + |
| 156 | +<p align="center"> |
| 157 | + <img src="images/extensions.png" alt="Extensions" width="600" /> |
| 158 | +</p> |
| 159 | + |
| 160 | +--- |
| 161 | + |
| 162 | +### π Browser Automation |
| 163 | +Two backends to choose from: |
| 164 | + |
| 165 | +| Backend | Best For | |
| 166 | +|---------|----------| |
| 167 | +| **Playwright** | Simple scraping, headless automation, zero setup | |
| 168 | +| **BrowserOS** | Real browser with your extensions, cookies, and logins | |
| 169 | + |
| 170 | +BrowserOS lets the AI control YOUR actual browser β including logged-in sessions, AdBlock, 1Password, and everything else. |
| 171 | + |
| 172 | +--- |
| 173 | + |
| 174 | +### π§ Persistent Memory |
| 175 | +OpenWhale remembers everything, even after restarts. |
| 176 | + |
| 177 | +- **Long-term Memory** β Facts and preferences in MEMORY.md |
| 178 | +- **Daily Notes** β Automatic daily logging |
| 179 | +- **Vector Search** β Find related content by meaning (local embeddings, no API needed!) |
| 180 | +- **Session Persistence** β Continue conversations across restarts |
| 181 | + |
| 182 | +--- |
| 183 | + |
| 184 | +### π Enterprise Security |
| 185 | +Production-ready security out of the box. |
| 186 | + |
| 187 | +- JWT authentication with session management |
| 188 | +- API key protection |
| 189 | +- Rate limiting |
| 190 | +- Audit logs |
| 191 | +- Multi-user support |
| 192 | +- Sandboxed code execution |
| 193 | + |
| 194 | +--- |
34 | 195 |
|
35 | 196 | ## Why OpenWhale? |
36 | 197 |
|
37 | | -- **It's agentic** β not just a chatbot. It can use tools, run commands, send messages, and interact with the real world. |
38 | | -- **Use any model** β Claude, GPT-4, DeepSeek, Groq, Gemini, Ollama. Switch whenever you want. |
39 | | -- **Runs anywhere** β Docker, bare metal, your Raspberry Pi. Whatever works for you. |
40 | | -- **Actually secure** β JWT auth, API keys, rate limiting, audit logs. The boring stuff that matters. |
| 198 | +| Feature | OpenWhale | ChatGPT | Claude | Typical Chatbot | |
| 199 | +|---------|-----------|---------|--------|-----------------| |
| 200 | +| **Multi-model support** | β
8 providers | β GPT only | β Claude only | β Single model | |
| 201 | +| **Run shell commands** | β
Full access | β No | β No | β No | |
| 202 | +| **Browser automation** | β
Playwright + BrowserOS | β Limited | β
Limited | β No | |
| 203 | +| **WhatsApp/Telegram/Discord** | β
All channels | β No | β No | β Maybe one | |
| 204 | +| **Twitter integration** | β
Full API | β No | β No | β No | |
| 205 | +| **Self-extensible** | β
Creates own tools | β No | β No | β No | |
| 206 | +| **Runs locally** | β
Your machine | β Cloud only | β Cloud only | β Usually cloud | |
| 207 | +| **Open source** | β
MIT license | β Closed | β Closed | β Varies | |
| 208 | +| **Persistent memory** | β
Vector search | β
Limited | β
Limited | β Usually no | |
41 | 209 |
|
42 | 210 | --- |
43 | 211 |
|
|
0 commit comments