OpenClaw vs Hermes Agent
OpenClaw (OpenClaw Foundation (nonprofit); created by Peter Steinberger) and Hermes Agent (Nous Research) side by side on public data, as of Nov 20, 2026.
OpenClaw has 35.4x the weekly package downloads of Hermes Agent (3.1M against 88.2K). OpenClaw has 1.5x the GitHub stars of Hermes Agent (395K against 265K). OpenClaw shipped 37 stable releases in the past 90 days. Hermes Agent shipped 20 stable releases in the past 90 days.
Which should you choose?
Choose OpenClaw if
- you need to run local models
- you want to switch between model providers (12 documented against 3)
- you want to check in from your phone
- you want frequent updates (37 stable releases in 90 days against 20)
Built from documented features, official prices, status pages and usage data on this page; it does not judge output quality.
Side-by-side numbers
| Metric | OpenClaw | Hermes Agent |
|---|---|---|
| Gid Score | 71 | 63 |
| Usage data as of | Nov 19 | Nov 19 |
| Rank by Gid Score | #8 | #13 |
| Downloads, last 7 days | 3.1M | 88.2K |
| Downloads, 7-day change | up 5.8% | up 34.8% |
| Downloads, 30-day change | down 24.3% | — |
| VS Code installs | — | — |
| Marketplace rating | 3.5 / 5 (78) | — |
| GitHub stars | 395K | 265K |
| Homebrew installs, 30 days | 1,317 | 2,185 |
| Hacker News mentions, 30 days | 123 | 46 |
| Latest stable release | 2026.10.33 · 2d ago | 0.23.2 · 4d ago |
| Stable releases, 90 days | 37 | 20 |
| Terminal-Bench (best run) | — | — |
| Price | Free (OSS) | Free (OSS) |
| License | MIT | MIT |
| First released | November 2025 | February 2026 |
| Interfaces | Messaging apps, CLI, Desktop, Mobile | CLI, Messaging apps, Desktop |
| Models | Model-agnostic: Claude, OpenAI Codex/GPT, DeepSeek and local models via plugins | Any model: Nous Portal, OpenRouter, OpenAI and custom endpoints; a Claude Pro or Max plan through an experimental plugin |
A dash means there is no public data for that metric as of Nov 20, 2026.
Trends
OpenClaw
Open-source personal AI assistant that runs on the user's own machine as a gateway and works through chat apps such as WhatsApp, Telegram, Slack, Discord and iMessage, with a skills system and swappable model plugins. It also ships native apps for macOS, iOS, Android, Windows and Linux. Its README warns that tools run on the host for the main session unless sandboxing is configured, and that inbound messages should be treated as untrusted input.
Pricing: Free and open source (MIT) with no paid tier, hosted service or token; users pay their chosen model providers.
Latest release 2026.10.33:
- Long conversations and large session stores use bounded export and recovery work, reducing out-of-memory failures during chat, memory export, Doctor repair, and Gateway startup. Authored
- Interrupted updates recover when their executor dies instead of leaving the Gateway stuck offline, while version-skew checks protect state still used by an installed service. Authored
- Replies survive more restart and reconnect paths, preserving original prompts, recovering Telegram responses, and avoiding lost WhatsApp replies or skipped Signal messages.
- Shared-session privacy is stronger: Side chat stays private to its viewer, internal model prompts stay out of displayed output, and compaction redacts credentials in quoted summaries. Authored
Hermes Agent
Open-source agent from Nous Research, described by its makers as self-improving, that keeps persistent memory, creates and refines skills from experience, runs scheduled automations, and works from a terminal interface or messaging apps such as Telegram, Discord, Slack and WhatsApp. It can run locally or on Docker, SSH, Modal, Daytona and Vercel Sandbox backends, and has a desktop app.
Pricing: Free and open source (MIT); users pay their chosen model providers (Nous Portal, OpenRouter, OpenAI or custom endpoints).
Latest release 0.23.2:
- Desktop chats created by an August build with a bare custom provider can send again after updating.
- A Desktop or TUI reply that streamed beside a todo or memory call now settles instead of leaving the Stop button up.
- A crash mid-stream on Telegram and other edit-streaming chats no longer shows a duplicate "Recovered reply"; a word holding a key prefix mid-word (program_id) is shown on time again.
- Desktop no longer loses or scrambles what you type while it starts and resumes your last chat.
OpenClaw vs Hermes Agent: common questions
Which is more popular, OpenClaw or Hermes Agent?
OpenClaw has a Gid Score of 71 against 63 for Hermes Agent. The score combines package downloads, editor installs, GitHub activity and attention, updated daily.
Do OpenClaw and Hermes Agent support local models?
OpenClaw documents support for local models; Hermes Agent's documentation does not.