Free alternatives to AgentOS
200 free apps like AgentOS, closest match first.
Top picks
- O

Open Multi-Agent
LaunchOpen Multi-Agent is a self-hosted TypeScript runtime for teams of AI agents, released under the MIT licence. You can start from a goal and let the runtime plan a directed graph of tasks, or write an explicit task graph yourself. A team can
Self-hosted
- S

Statewave
LaunchStatewave is an open source memory runtime for AI agents. Instead of retrieving chunks at query time, it records raw events as immutable episodes, compiles them into typed semantic and episodic memories with provenance, and returns ranked,
Self-hosted
- C

Cognithor
LaunchCognithor · Agent OS: Local-first autonomous agent operating system. 19 LLM providers, 18 channels, 145 MCP tools, 6-tier memory, Agent Packs marketplace, zero telemetry. Python 3.12+, Apache 2.0.
- F
feros
LaunchOpen-source voice agent OS. Rust runtime, AI-driven builder, sub second latency. Self-host everything.
Self-hosted
- O
OS
LaunchEnables organizations to create and deploy customizable AI agents with support for multiple LLM providers, real-time chat, voice, document training, analytics, user management, SSO authentication, and embeddable widgets. Built with React 19
Self-hosted
- A
Agent Memory
Launchagent-memory is a long-term memory runtime for AI agents. It keeps plain Markdown files in one store as the source of truth, with a SQLite index beside them that is a cache and can be deleted at any time. Claude Code, Codex CLI and any othe
More like AgentOS
About AgentOS
AgentOS
Open-source TypeScript runtime for autonomous AI agents
AgentOS is an open-source TypeScript runtime and framework for building autonomous AI agents, licensed under Apache 2.0 and made by Frame.dev. It covers multi-agent orchestration, cognitive memory with an Ebbinghaus forgetting curve, multimodal retrieval-augmented generation, optional safety guardrails, voice pipelines and support for 13 LLM providers, including OpenAI, Anthropic, Google Gemini, Groq, Mistral and Ollama. Agents can write their own functions mid-task, with a judge approving each one, and managers can spawn specialist agents at runtime when the static roster does not cover a task. Each session on the full runtime is a Generalized Mind Instance that keeps its own persona, mood and memory. The site reports 85.6% on the LongMemEval-S benchmark and compares the framework with LangGraph, CrewAI, Mastra and AutoGen. It installs with npm and needs Node.js 22 or higher.
- Price
- Open source
- Category
- AI
- Website
- agentos.sh







