Product / UX / Orchestration & agents

Hermes Agent

Open-source autonomous agent that grows with the user through persistent memory, reusable skills, and a built-in learning loop. It runs on a VPS or serverless backend, can be controlled through Telegram and other messaging surfaces, and is designed to keep improving over long-lived sessions rather than acting as a one-shot chat wrapper. Scale signal: the Hermes Agent repo is MIT-licensed, has 163k+ GitHub stars, and shows very heavy recent development activity.

Hermes Agent by Nous Research, open source self-improving autonomous agent (auto-generates skills, persistent memory, 40+ tools, subagent delegation, NL cron, multi-channel). MIT. Released ~Feb 2026; verified live June 2026 (latest release v0.15.2, 29 May 2026; ~181K stars).

Openness

5 high confidence

5.0

license: MIT(OSI)
source: public(github.com/NousResearch/hermes-agent)
core-gated: ungated

Fully MIT-licensed, complete public source, self-hosted by design, no proprietary core; top openness, above open-core framework anchors.

https://github.com/NousResearch/hermes-agent recorded 2026-06-04
MIT LICENSE, full public source, self-improving multi-channel agent (Python)
https://hermes-agent.nousresearch.com recorded 2026-06-04
open source MIT autonomous agent product page

Adoption

4 medium confidence

4.0

Real usage signal beyond stars (adoption!=stars rule): Hermes Agent reported #1 on OpenRouter's GLOBAL token rankings, confirmed by Nous Research's own announcement (~271B tokens, May 6 2026) and an OpenRouter app page; secondary reports give ~224B daily tokens by May 10 (surpassing OpenClaw's ~186B). Both figures are reported; token-volume mapped to 1-10M-equivalent reach (inference, not a disclosed installed-user count). GitHub stars (~181K, <4 months) corroborate but are not the basis. Level 4 / medium confidence: deciding metric now backed by a primary source, but reach is inferred from token volume.

https://github.com/NousResearch/hermes-agent recorded 2026-06-04
active releases (latest v0.15.2, May 29 2026), ~181K stars in <4 months (corroborating, not basis)
https://x.com/NousResearch/status/2052904761087729897 recorded 2026-06-04
PRIMARY: Nous Research announcing Hermes Agent #1 on OpenRouter global token rankings
https://openrouter.ai/apps/hermes-agent recorded 2026-06-04
PRIMARY: OpenRouter app page for Hermes Agent (token-usage ranking platform)
https://chatforest.com/builders-log/hermes-agent-nous-research-openrouter-number-one-self-improving-builder-guide/ recorded 2026-06-04
aggregator detail: ~271B tokens, surpassing OpenClaw's 186B (corroborating)

Capability

4 medium confidence

4.0

Distinctive autonomy (self-improvement loop + persistent memory + subagent delegation) and very broad tool/channel coverage place it high on the feature matrix; no published benchmark so scored 4, just below a verified-benchmark ceiling.

https://github.com/NousResearch/hermes-agent recorded 2026-06-04
self-improving skills, persistent memory, 40+ tools, subagents, cron, sandboxing features

Unchanged since 2026-07-30 (last edited, not re-checked)