Auto Learning Agents
Overview
Auto Learning Agents is a free, open-source AI agent platform that can be self-hosted within minutes using Docker. Users can clone the repository, add their API keys and run their own AI agents.
It's compatible with a variety of models, such as Claude, OpenAI, or Gemini, and users can mix these models according to their requirements. It also supports fully local operations with tools like Ollama and without any external keys.
Auto Learning Agents comes bundled with Elixir runtime, Python services, a local database, and a full tool layer, encapsulated within a single Docker image.
It's designed to be easily installed with just a few commands following the installation of Docker and Docker Compose. The platform is designed with light requirements.
Docker is the only requirement on the user's machine and at least one API key from a provider like Anthropic, OpenAI, or Google, or a local model served through Ollama, for the agents to operate.
It leverages a single bundled database for storage and search, avoiding the need for any external setup or a cloud account. The platform can be used for a wide range of applications, including browser automation, fault tolerance, code review, and more.
Comprehensive documentation is provided, covering configuration, model selection, memory backend, and an in-depth understanding of the system behind the platform.
Releases
Top alternatives
-
AI Agents hallucinate. Catch it. Fix it fast.
-
Deploy your private AI agent in minutes.
Philip Alfredsson🙏 1 karmaMay 11, 2026@Hermes Agent HelperHermes Agent Helper seems genuinely useful for people experimenting with AI agents without wanting to deal with overly complex setups. I like that the product feels more practical and workflow-focused instead of just AI hype. -
The engine for AI agent analytics and simulation.
-
AI agents that actually run your company.Hello, If you had to describe your AI in one sentence without using the word "AI," what would you say? The reason I asked is because it's an interesting way to understand what makes a product truly valuable. How would you describe it?
-
The managed control plane for your AI agents.
-
Capability layer for the agentic economy.

