ZeroClaw

ZeroClaw, an open-source AI assistant from zeroclaw-labs, has reached 32,933 GitHub stars with version v0.8.5 released under a dual MIT OR Apache-2.0 license. The Rust-built binary starts in under 20ms, runs on ARM and x86 hardware including Raspberry Pi, and connects to 70+ LLM providers and 30+ channels while enforcing OS-level sandboxes via Landlock, Bubblewrap, Seatbelt, or Docker. The project positions itself as a self-hosted alternative to cloud AI assistants, using the user's own API keys or fully local models such as Ollama.

Open Source Alternative to: 32,933stars+57 +0.2% Last 30 days - Last commit 15 hours ago - Repository age 8 months - Version v0.8.5 - License Apache-2.0 https://openalternative.co/licenses/apache-2 - Self-hosted Yes https://openalternative.co/collections/self-hosted - Repository zeroclaw-labs/zeroclaw https://github.com/zeroclaw-labs/zeroclaw Most AI assistants are a seat you rent on someone else's infrastructure. ZeroClaw runs as a single native binary on your own machine, using your own API keys or fully local models like Ollama https://openalternative.co/ollama . No cloud seat, no subscription, no ZeroClaw server in the middle. The binary is built in Rust and starts in under 20ms. It uses less memory than a browser tab and runs on everything from a workstation to a Raspberry Pi, including ARM and x86. There's no Node, JVM, or Python environment to install. What it connects to: - 70+ LLM providers: local options like LM Studio https://openalternative.co/alternatives/lm-studio , llama.cpp, vLLM, and LocalAI https://openalternative.co/localai , plus hosted providers including Anthropic, OpenAI, Google Gemini, Amazon Bedrock, Mistral, Groq, and any OpenAI-compatible endpoint - 30+ channels: Telegram, Discord, WhatsApp, Slack, Signal, iMessage, Matrix, email, IRC, Bluesky, Reddit, Nostr, Notion, webhooks, and more - Hardware: GPIO on Raspberry Pi, STM32, Arduino, and ESP32 Security is the default posture, not an add-on. - Supervised autonomy: medium-risk actions require your approval before they run; high-risk ones are blocked outright - OS-level sandboxes: Landlock, Bubblewrap, Seatbelt, or Docker contain what the agent can touch, enforced by the kernel rather than a prompt - Command allowlists: explicit workspace scoping decides what runs and where - Tool receipts: every successful tool call can be stamped with an HMAC-SHA256 tag the model cannot forge, making fabricated runs or invented results detectable YOLO mode exists for trusted dev environments and is strictly opt-in. The agent supports multiple named agents running from a single daemon, each isolated. Skills package repeatable tasks with their tools. Scheduled jobs, webhook triggers, and channel events all run through the same agent loop inside the same sandbox and allowlists you've scoped. A2A discovery lets agents describe and find one another through the gateway. With a local model, nothing leaves your machine at all. With a hosted provider, only your prompts go to the provider you chose, using your own key. Dual-licensed MIT OR Apache-2.0. Every Sunday we deconstruct one proprietary app and pick the best open source alternatives worth switching to. - Stars - Last commit - License AI An AI assistant that acts, not just answers - Stars 391,408 - Last commit 6 hours ago - License MIT A personal AI assistant that connects to your existing chat apps and handles real tasks like email, calendar, and flight check-ins on your behalf. Compare with Desktop AI agent with persistent memory across platforms - Stars 251,285 Build and run AI agents that handle recurring work for you - Stars 187,658 Manage, build, and run AI agent teams around the clock - Stars 82,991 Personal AI agent for messaging apps, containerized - Stars 30,875 AI workforce platform where agents, skills, and memory live in your repo - Stars 20,250 Agentic AI framework that runs, learns, and self-corrects - Stars 19,376 Self-hosted enterprise agent platform with real permissions - Stars 6,049 Self-hosted AI agents that do real work on your machine Last commit 6 hours ago License MIT Cross-platform desktop agent from Nous Research that connects to Telegram, Discord, Slack, WhatsApp, and more, with persistent memory, scheduling, and isolated sandboxing. 8 hours ago Other Platform for building, deploying, and running AI agents that automate repeatable digital work across research, outreach, content, and support without writing code. Unknown A collaborative platform to create, schedule, and operate AI agents that handle long-running tasks, team workflows, and automated jobs without constant oversight. Self-hosted AI agent that connects to WhatsApp, Telegram, Slack, and a dozen more messaging apps, running each agent in an isolated Docker container with credential injection via a secure vault. 7 hours ago Elastic-2.0 Manages a team of AI agents across your company, with every agent, skill, memory, and connector stored as versioned files in a git repo you own and self-host. 3 days ago Autonomous AI agent framework that operates on its own sandboxed computer, creates tools dynamically, manages memory, and executes multi-step workflows with full transparency. Deploy AI agents inside your own infrastructure with document-grounded answers, computer use, and support for any AG-UI-compatible agent framework. Stars 3,328 9 hours ago Apache-2.0 Run persistent AI bots that sign into your tools, complete tasks autonomously, and ask for approval when needed. Your keys, your model, your infrastructure.