Cline
An open-source agent in your editor where you control the model and pay only for inference.
verified 2026-11-03
Agent harnesses (open source) · open source
Hermes Agent
Nous Research's open-source agent harness. It has a terminal UI and a messaging gateway (Telegram, Discord, Slack, WhatsApp, Signal), 40+ tools, persistent memory, skills it creates and improves from experience, a cron scheduler and subagents. It runs on a laptop, a $5 VPS or a serverless backend.
Best for: A self-hosted personal agent with long-term memory and self-written skills that you can reach from the terminal or from Telegram, Discord or Slack.
Harness specs
| Models & providers | Any model through 40+ built-in providers: Nous Portal, OpenRouter, Anthropic (API key, or Claude Max extra-usage credits via OAuth), OpenAI API or ChatGPT/Codex sign-in, Google Gemini and Vertex AI, xAI, DeepSeek, AWS Bedrock, Azure AI Foundry, Hugging Face, Kimi, MiniMax, z.ai GLM and more. Switch with hermes model. |
|---|---|
| Runs local / offline | yes Docs cover fully local setups with Ollama, vLLM and llama.cpp, plus LM Studio as a built-in provider, all through an OpenAI-compatible endpoint (for example http://localhost:11434/v1 for Ollama). The local model must serve at least 64K tokens of context. |
| License | MIT |
| Setup | Low. One-line installer (curl … | bash on Linux, macOS and WSL2; PowerShell on Windows), then hermes setup or hermes model to pick a provider. There's also a Termux package for Android. |
| Repo | github.com/NousResearch/hermes-agent |
| Sources | [1] [2] [3] |
Pricing
| Free tier | Yes. Free and open source (MIT); you pay only for the model provider you connect. |
|---|---|
| Paid plans | Optional Nous Portal subscription covers 300+ models plus a Tool Gateway (web search, image generation, TTS, cloud browser) under one login. The Portal price isn't listed in the official docs, so check portal.nousresearch.com. |
| Maker | Nous Research |
| Source | https://github.com/NousResearch/hermes-agent |
| Last verified |
Pros
- Built-in learning loop: agent-curated memory, self-improving skills and cross-session search
- One gateway process serves the CLI and Telegram, Discord, Slack, WhatsApp and Signal
- Six terminal backends: local, Docker, SSH, Modal, Daytona and Vercel Sandbox
Cons
- Local models need at least a 64K-token context window, and smaller windows are rejected at startup
- Not a coding-first tool, so it's broader and heavier than a pure code agent
Notable facts
- Docs list 40+ inference providers, including Nous Portal, OpenRouter, Anthropic, OpenAI, Gemini, Bedrock, LM Studio and any custom OpenAI-compatible endpoint — source
- Can import settings, memories and skills from OpenClaw, Claude Code, Codex CLI and Cursor — source
Alternatives
Aider
Terminal pair programming with automatic git commits and any LLM, including local models.
verified 2026-11-03
OpenCode
Trying a capable open-source terminal agent for $0, with free models and no account.
verified 2026-11-03
Goose
An extensible open-source agent that can install, execute, edit and test with any LLM.
verified 2026-11-03
OpenHands
Self-hosting a team of coding agents, sandboxed in Docker, with scheduled or webhook-triggered automations.
verified 2026-11-03
FAQ
Is Hermes Agent free?
Yes. Free and open source (MIT); you pay only for the model provider you connect.
How much does Hermes Agent cost?
Optional Nous Portal subscription covers 300+ models plus a Tool Gateway (web search, image generation, TTS, cloud browser) under one login. The Portal price isn't listed in the official docs, so check portal.nousresearch.com.
What is Hermes Agent best for?
A self-hosted personal agent with long-term memory and self-written skills that you can reach from the terminal or from Telegram, Discord or Slack.
Who makes Hermes Agent?
Hermes Agent is made by Nous Research.
Spotted something out of date? Tell us and we'll re-verify. Facts on this page come only from the linked sources.