Hermes — The self-improving AI agent

+

The core idea

+

Why Hermes?

+

+ Most AI tools are excellent in the moment and weak over time. They give you a great answer, then forget you exist. The next session starts from zero. +

+

+ The distinction isn't "has memory" vs "no memory" anymore. Every major tool has added some form of memory in the past year. The real question is whether that context is durable (survives reboots), self-hosted (your hardware), cross-surface (same agent on Telegram, Slack, terminal), and cumulative (compounds over months, not resets per project). +

+

+ Hermes answers yes to all four. It runs as a persistent process on your server. Memory is markdown files in ~/.hermes/. The same agent that answered your Telegram message at 9am is available in your terminal at 2pm, with full context. +

+

+ The market is converging fast. Chat assistants have added scheduling. Editors are shipping cloud agents. CLIs are getting skills systems. Hermes wins through synthesis — all of these capabilities working together on hardware you own, with a model of your choice. +

+

+ Why synthesis matters: A scheduler without messaging is just cron. Messaging without memory is just a chatbot. Memory without scheduling is a glorified notes app. Hermes ties them together into a single persistent process with an identity that accumulates across time. +

+

Session-scoped tool

+

+ You → [Tool] → Answer → Done +

+

Persistent agent

+

+ You ↔ [Hermes] ↔ (memory, skills, +

+

+ schedule, tools, +

+

+ surfaces) +

+

+ Memory: ~/.hermes/ (markdown files)
+ Skills: auto-written from experience
+ Schedule: cron, runs while you sleep
+ Surfaces: Telegram · Discord · Slack · browser
+ Model: your choice, swap anytime +

+

Three pillars

+

What makes it different

+

+ +

+

🧠

+

Memory that compounds

+

Layered memory system: user profile, agent memory, skills, and session history. All stored locally as markdown files at ~/.hermes/.

+

Survives every reboot and model swap
8 optional external memory providers
Grows more useful the longer you use it
Portable — it's just files

+

+ + Memory docs → + +

+ +

+

⏰

+

Autonomous scheduling

+

Built-in cron scheduler runs on your server and delivers results to any platform you've connected. Set it and forget it.

+

Morning news briefings to Telegram
PR review automation
Test suite monitoring
Blog watchers and price alerts

+

+ + Scheduling docs → + +

+ +

+

📡

+

Reach it from anywhere

+

15+ messaging platforms, same agent, same memory. Switch surfaces mid-conversation without losing context.

+

Telegram · Discord · Slack · WhatsApp
Signal · Matrix · Mattermost · Email · SMS
DingTalk · Feishu · WeCom · BlueBubbles
Home Assistant · browser

+

+ + Messaging docs → + +

+ +

+

Full feature set

+

Everything in one system

+

+ +

+

🔧

+

47 built-in tools

+

Web search and extraction, browser automation, code execution, vision analysis, image generation, TTS, subagent delegation, and more.

+ Tools reference → +

+ +

+

📚

+

Self-improving skills

+

The agent writes its own skills from experience. Compatible with agentskills.io open standard and shareable via Skills Hub.

+ Skills docs → +

+ +

+

🔌

+

MCP integration

+

Connect to any MCP server — GitHub, databases, file systems, APIs. Hermes can also expose itself as an MCP server for Claude Code, Cursor, or Codex.

+ MCP docs → +

+ +

+

🎙️

+

Voice mode

+

Real-time voice in CLI (Ctrl+B to record), Telegram voice bubbles, Discord voice channels. Supports faster-whisper locally or Groq/OpenAI Whisper.

+ Voice docs → +

+ +

+

🖥️

+

6 terminal backends

+

Local, Docker, SSH, Daytona, Singularity, Modal. Run execution anywhere from a $5 VPS to serverless cloud, sandboxed as needed.

+ Backend docs → +

+ +

+

🎭

+

Personality & SOUL.md

+

Define the agent's default voice with a global SOUL.md file. 14 built-in personas plus custom personalities via config.

+ Personality docs → +

+ +

+

🤝

+

Agent orchestration

+

Spawn Claude Code or Codex as sub-agents for heavy coding tasks. Results fold back into Hermes memory. Hermes also runs as an MCP server for other tools.

+ Orchestration docs → +

+ +

+

🔒

+

Security model

+

7-layer defense in depth: user allowlists, dangerous command approval, Docker isolation, MCP credential filtering, prompt injection scanning, cross-session isolation, input sanitization.

+ Security docs → +

+ +

+

Hermes Web UI

+

A browser interface for Hermes

+

+ Full-featured, no build step, no framework. Three-panel Claude-style layout — sessions sidebar, chat, workspace file browser. + GitHub → +

+ +

+

What it includes

+

1:1 parity with Hermes CLI

+

Multi-profile support

+

Session history with projects and tags

+

Tool call cards with expand/collapse

+

Workspace file browser

+

Inline file preview

+

Light + dark mode

+

Password protection

+

SSH tunnel ready

+

Docker image available

+

+ +

+

1

+

Install Hermes Agent first

+

+ bash + +

+

curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash
+hermes model   # configure your LLM provider

+

+ +

+

2

+

Clone and start the Web UI

+

+ bash + +

+

git clone https://github.com/nesquena/hermes-webui.git hermes-webui
+cd hermes-webui
+./start.sh

+

+ +

+

or

+

Docker option

+

+ bash + +

+

docker pull ghcr.io/nesquena/hermes-webui:latest
+docker run -d \
+  -e WANTED_UID=$(id -u) -e WANTED_GID=$(id -g) \
+  -v ~/.hermes:/home/hermeswebui/.hermes \
+  -v ~/workspace:/workspace \
+  -p 8787:8787 ghcr.io/nesquena/hermes-webui:latest

+

+ +

+

→

+

SSH tunnel (for remote servers)

+

+ bash + +

+

ssh -N -L 8787:127.0.0.1:8787 user@your-server

+

+ Then visit http://localhost:8787 +

+

Honest comparison

+

How Hermes compares

+

+ The agent landscape is converging fast. Chat assistants added scheduling. Editors shipped cloud agents. CLIs are getting skills systems. Most comparisons a year ago are already outdated. This table reflects where things stand in mid-2026 — and will keep shifting. Hermes wins through synthesis: these capabilities working together on hardware you own, with a model of your choice. +

+ +

vs. the agent landscape

+

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

Tool	Persistent memory	Self-hosted scheduling	Messaging / surfaces	Self-hosted	Open source	Self-improving skills	Always-on
Hermes	✓	✓ self-hosted	✓ 15+ platforms	✓	✓ MIT	✓ automatic	✓
OpenClaw	✓	✓	✓ 24+ platforms	✓	✓ MIT	Partial	✓
Claude Code	Partial	Cloud only	Preview	No	No	No	No
Codex	Partial	Desktop app	No	✓ (CLI)	✓ Apache 2.0	No	No
OpenCode	Partial	No	Community	✓	✓	Community	No
Cursor	✓ per-project	Cloud VMs	✓ Slack/web	No	No	No	No
GitHub Copilot	Partial repo	Via agent	Via CLI	No	No	No	No
Claude.ai	✓	✓ Cowork	✓ 50+ connectors	No	No	No	No
ChatGPT	✓	✓	✓ 50+ connectors	No	No	No	No

+

+ +

Head-to-head

+

+ +

+

Hermes vs. OpenClaw

+

+ OpenClaw (MIT, ~350k+ GitHub stars) is the closest comparison — open-source, self-hosted, always-on, with persistent memory, cron scheduling, and messaging. OpenClaw has the widest messaging surface (24+ platforms, including iMessage, LINE, WeChat, Teams) and native Chrome CDP browser control. Hermes wins on stability (OpenClaw has documented Telegram regressions through April 2026) and security — three separate 2026 audits found malicious skills in ClawHub, with Koi Security linking 335 packages to the "ClawHavoc" campaign and Bitdefender flagging ~900 malicious packages (~20% of the ecosystem). Hermes writes skills automatically; OpenClaw centers on a human-curated marketplace. Hermes also runs in the Python/ML ecosystem; OpenClaw is Node.js. +

+

+ +

+

Hermes vs. Claude Code

+

+ Claude Code is Anthropic's official agentic tool, strong for focused coding sessions. It has a 26-event hooks system, a plugin/skills marketplace, CLAUDE.md/MEMORY.md project memory, and cloud-managed scheduling on Anthropic infrastructure. Research-preview messaging covers Telegram, Discord, and iMessage; Slack is not yet shipped. The key differences: scheduling runs on Anthropic's cloud (your data leaves your hardware), it's locked to Claude models, and it's not open source. Hermes can spawn Claude Code as a sub-agent for heavy implementation tasks — they're complementary. +

+

+ +

+

Hermes vs. Cursor

+

+ Cursor has changed substantially. Per-project Memories shipped June 2025. Automations launched March 2026 — time-based, event-based, Slack-triggered background agents. v3.0 (April 2026) went agent-first. It's now valued at $29.3B. Cursor is excellent at what it does. But its agents run on cloud VMs with no data sovereignty. Hermes runs on your server, accumulates context across all your projects (not per-project), works with any model family, and is reachable from every surface you use — not just the editor. +

+

+ +

+

Hermes vs. Claude.ai / ChatGPT

+

+ Both are capable and no longer just chat tools. Claude Cowork added scheduling (Feb 2026) and 50+ connectors including Slack. ChatGPT has Agent Mode (July 2025), Scheduled Tasks, a computer-using agent, and 50+ connectors. The fundamental issue isn't features — it's where execution lives. Neither is self-hosted. Memory, sessions, and scheduled tasks run on their servers. For workflows requiring data sovereignty, server-resident execution, or provider flexibility, that's a hard constraint with no workaround. +

+

+ +

+

Quick start

+

Get started in 60 seconds

+ +

+

+ Works on Linux, macOS, WSL2, and Android (Termux). On Windows, use WSL2 first. +

+ +

+

+ bash — one-liner install + +

+

curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash

+

+ +

+

+ bash — first run + +

+

source ~/.bashrc      # reload shell
+hermes model          # choose your LLM provider
+hermes                # start chatting

+

+ +

+

+ bash — key commands + +

+

hermes model          # choose LLM provider and model
+hermes tools          # configure enabled tools
+hermes gateway setup  # set up messaging platforms
+hermes doctor         # diagnose any issues

+

+ + + Full installation guide → + +

+ +

+

Supported LLM providers

+

+ Nous Portal + OAuth +

+

+ OpenAI / Codex + Device code auth +

+

+ Anthropic + API key / auth +

+

+ OpenRouter + 200+ models + API key +

+

+ GitHub Copilot + OAuth or token +

+

+ DeepSeek + API key +

+

+ Groq + API key +

+

+ Hugging Face + 20+ models + API key +

+

+ Ollama / local + Local +

+

+ Custom endpoint + Any OpenAI-compatible URL + Custom +

+

+ ⚡ + Run hermes model to browse and authenticate interactively +

+

Honest limitations

+

What Hermes is not

+

+ Worth saying clearly. These are real trade-offs, not caveats. +

+

+ +

+

✏️

+

+ Not the best in-editor autocomplete. For tight editor integration and line-by-line completions, use Cursor or Windsurf. That's not what Hermes is optimizing for. +

+

+ +

+

🔧

+

+ Not zero-setup. You're running a server process. There's initial configuration for providers, messaging platforms, and tools. That's the trade-off for having something you own. +

+

+ +

+

🤖

+

+ Doesn't make weaker models magical. The infrastructure is only as good as the model you connect. A weak model with great memory is still a weak model. +

+

+ +

+

⚠️

+

+ Needs guardrails for high-stakes automations. Autonomous scheduling with broad tool access is powerful. Review what you're scheduling. The security model helps, but judgment matters. +

+

+ +

+

⚡

+

+ Not the lowest-friction one-shot tool. For a quick answer you'll never revisit, claude.ai or ChatGPT is faster to reach. Hermes earns its setup cost over time, not in the first session. +

+

+ +

+

Docs & community

+