Hermes Agent, developed by Nous Research, has become one of GitHub’s fastest-growing AI projects, earning nearly 6,000 stars in a single day. The project represents a new approach to AI agents: one that learns and improves over time rather than starting fresh each session.
The Agent That Grows With You
Hermes is what Nous Research calls the agent that grows with you. It is the only AI agent with a built-in learning loop. It creates skills from experience, improves them during use, nudges itself to persist knowledge, searches its own past conversations, and builds a deepening model of who you are across sessions.
The self-improving AI agent built by Nous Research runs on a remarkable range of infrastructure. You can deploy it on a 5 dollar VPS, a GPU cluster, or serverless infrastructure that costs nearly nothing when idle. It is not tied to your laptop. You can talk to it from Telegram while it works on a cloud VM.
Model Flexibility Without Lock-In
One of Hermes’s key strengths is its model agnosticism. You can use any model you want: Nous Portal, OpenRouter with access to 200+ models, GLM, Kimi/Moonshot, MiniMax, OpenAI, or your own endpoint. Switching models is as simple as running hermes model. No code changes, no lock-in.
This flexibility extends to deployment options as well. Six terminal backends are supported: local, Docker, SSH, Daytona, Singularity, and Modal. Daytona and Modal offer serverless persistence where your agent’s environment hibernates when idle and wakes on demand.
Memory and Learning That Persist
Traditional AI assistants treat every conversation as isolated. Hermes breaks this pattern with agent-curated memory that includes periodic nudges for self-improvement. After complex tasks, the agent autonomously creates new skills that it then improves during use.
The system includes FTS5 session search with LLM summarization for cross-session recall, and it is compatible with the Honcho dialectic user modeling system. The skills system allows Hermes to build procedural memory that shapes future interactions.
Messaging Across Platforms
Hermes lives where you do. Through its gateway system, you can interact with it via Telegram, Discord, Slack, WhatsApp, Signal, and CLI, all from a single gateway process. The system supports voice memo transcription and cross-platform conversation continuity.
The built-in cron scheduler allows for scheduled automations with delivery to any platform. Daily reports, nightly backups, weekly audits, configured in natural language and running unattended.
Parallel Workstreams and Research Capabilities
For power users, Hermes can spawn isolated subagents for parallel workstreams and collapse multi-step pipelines into zero-context-cost turns via Python scripts that call tools via RPC. This makes it suitable for complex, multi-tasked workflows.
For researchers, Hermes includes batch trajectory generation, Atropos RL environments, and trajectory compression for training the next generation of tool-calling models.
Easy Installation
Getting started is straightforward. On Linux, macOS, or WSL2, installation is a single command: curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash
The installer handles everything, Python, Node.js, dependencies, and the hermes command. No prerequisites except git. Windows users are directed to install WSL2 first.
Open Source with Community Support
Released under the MIT license, Hermes welcomes contributions from the developer community. The project maintains active Discord and GitHub discussions, and it is compatible with the agentskills.io open standard.
For users coming from OpenClaw, Hermes offers automatic migration of settings, memories, skills, and API keys. The migration command hermes claw migrate can import SOUL.md, memories, skills, command allowlists, messaging settings, API keys, and TTS assets.
The Future of Personal AI
Hermes represents a shift in how we think about AI assistants. Rather than static tools that perform tasks and forget, Hermes agents are designed to accumulate knowledge, refine their approaches, and develop increasingly personalized understandings of their users over time.
As AI capabilities continue to advance, projects that enable persistent, self-improving agents like Hermes are likely to play an increasingly important role in how individuals and organizations harness these capabilities. With its combination of flexibility, power, and ease of use, Hermes is positioned at the forefront of this evolution.