agentos.guide › blog
Hermes Agent v0.22: The Quick Start, And How The Threshold Release Fits Your Agent OS
Hermes Agent v0.22 is the newest major release of Nous Research's free, open-source AI agent, and Nous calls it The Threshold Release.
The official release notes date it 13 October 2026, and a v0.22.1 patch followed on 14 October 2026.
A fresh desktop install now sets itself up through a chat on a free trial model, Hermes can work inside apps such as Linear and Notion, and local models install in one click.
This page is the quick start, plus how each new feature fits an Agent OS workflow.
What Hermes Agent v0.22 is, in one minute
Hermes Agent is the open-source AI agent from Nous Research that runs on your own machine, uses tools, learns skills and works through your chat apps.
Version 0.22 is the release Nous Research calls "The Threshold Release", because it is about getting people through the door.
| Fact | What the official release notes say |
|---|---|
| Version | Hermes Agent v0.22.0, which Nous Research calls "The Threshold Release". |
| Release date | The release notes are dated 13 October 2026. |
| Latest patch | v0.22.1 followed on 14 October 2026 with Windows Store fixes and catalog additions. |
| Maker and licence | Nous Research makes it, and it is free and open source under the MIT licence. |
| Size | About 24,200 commits, 6,080 merged PRs and 2,520 contributors went in since v0.21.0. |
| Codebase | The source shrank from about 1.06 million lines to about 698,000 lines. |
Nous Research says this is the largest window Hermes has shipped, at about 2.5 times the size of the v0.21.0 cycle.
In my video I grouped it into three moves, which were walk in, plug in and hand off.
Walk in covers the setup chat, the free welcome model, local models, signed apps and safer updates.
Plug in covers the plugin catalog, connectors and password-manager sign-in.
Hand off covers Bot Screen, reasoning that carries across sessions and waiting out provider slowdowns.
The video is a walkthrough of the release notes, so I am telling you what Nous Research says each feature does rather than showing a hands-on test of each one.
If you want the company background first, read my Hermes Agent by Nous Research guide.
The Hermes Agent v0.22 quick start
This is the shortest path from nothing to a working v0.22 setup.
| Step | What you do | Where |
|---|---|---|
| 1 | Update an existing install with hermes update, or let the desktop app update itself. | Terminal or desktop app |
| 2 | For a fresh install, run the official one-liner for your system. | Terminal or PowerShell |
| 3 | Open the desktop app and finish the setup chat, or click Skip setup. | Desktop app |
| 4 | Connect one app on the Connectors page. | Desktop app, /connectors or hermes connectors |
| 5 | Add a local model from Settings → Providers → Local Models. | Desktop app or hermes model |
| 6 | Install one plugin from the 600-plugin catalog. | Desktop app or CLI |
| 7 | Check your profiles on the single gateway with hermes gateway status. | Terminal |
For a fresh install on Linux, macOS or WSL2, the official one-liner is curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash.
On native Windows, you run iex (irm https://hermes-agent.nousresearch.com/install.ps1) in PowerShell.
Docker users pull nousresearch/hermes-agent:stable, and the notes say the stable and latest tags now move only when a release is published.
Microsoft Store installs update through the Store.
Source installs now follow tested SemVer releases by default, and hermes update --set-channel main puts you back on every commit if you really want that.
Read these upgrade notes before you click update
These are the changes most likely to surprise an existing Agent OS setup.
- Managed installs now run on Python 3.14, and the package manager installs it for you.
- The Windows desktop app is MSIX-only and needs Windows 11 22H2 or newer.
- The default macOS build supports Apple Silicon only.
- One messaging gateway per machine is now the default, and
hermes updatefolds per-profile gateways into it when nothing conflicts. - Chats no longer reset on an idle or daily timer, so
/newor/resetstarts a fresh chat. - Ended sessions inactive for 90 days are deleted at startup unless you set
sessions.auto_prune: false. - Memory providers such as Honcho, Supermemory, Hindsight and Mem0 moved to the plugin catalog, and existing users are migrated with data intact.
If your Agent OS keeps old Hermes chats as a record, set the auto-prune option first.
🔥 Want the exact Hermes Agent setup I run inside my Agent OS?
Inside the AI Profit Boardroom you get the installable Agent OS, step-by-step video tutorials, four coaching calls a week and 3,400+ members building real automations.
Walk in: the setup chat and the free welcome model
A fresh desktop install now opens on a guided setup chat instead of a settings screen.
It asks for your name, an accent colour and a layout, then which apps to connect and which plugins to install.
Then it gives a quick tour and offers up to four first tasks picked for you.
It runs on a free trial model called nous/welcome, once per operating-system user, so nobody needs a key to get started.
The agent also gained a set_theme tool, so "make it violet, dark mode" works in any desktop chat.
The catch is that the CLI, the TUI and the gateway get no free model, and existing installs and remote backends do not get the setup chat automatically.
When you are ready, hermes auth upgrade or /login turns the free identity into a full Nous account and keeps your connected apps.
Walk in: local models in one click
Hermes can now run open models entirely on your machine with no account and no API key.
It installs and maintains llama.cpp with the right engine for CPU, CUDA, Metal, Vulkan or ROCm.
It rates every catalog model against your GPU and RAM, and it only recommends models that fit fully in GPU memory.
You start from Settings → Providers → Local Models on Windows, macOS and Linux, or from hermes model then Local models in a terminal.
Downloads pause and resume, and the local server starts and stops with Hermes.
LM Studio now has its own provider profile, and a llama.cpp server you started yourself shows up in /model.
Plug in: connectors, plugins and password managers
Connectors let Hermes act inside hosted apps such as Linear and Notion through the Nous tool gateway.
You sign in once in your browser, and the tools work in the same turn.
Hermes creates a free connectors-only identity the first time, so you do not need a Nous account to start.
The new Connectors page lists hosted apps and your MCP servers together, and it replaces the MCP tab.
You can connect a work account and a home account for the same app and manage them with /connectors.
The plugin catalog grew to 600 entries this cycle, and each one is pinned to a reviewed commit.
You can browse it on the Hermes plugin docs under Memory, Desktop, Platforms, Web, Tools, Voice, Automation and Models.
My best Hermes Agent plugins guide is a good shortlist if 600 feels like too many.
The browser agent can now sign in with 1Password, Bitwarden or Hermes's own encrypted vault, and the model never sees the password.
One-time codes are filled without entering the conversation, and card and address fills need your confirmation.
Hand off: Bot Screen, subagents and outage waits
A bot running on a Linux gateway can now have its own desktop that Hermes Desktop streams live.
When it hits a login, a 2FA prompt or a CAPTCHA, you take the screen, finish that step and hand control back.
With a Docker, SSH or Singularity backend, the screen, computer use and browser run inside the sandbox.
The notes say Bot Screen is off by default and needs some host packages.
Running subagents now stay visible with their task, elapsed time and live transcript, plus Steer and Stop buttons.
You open that view with F6 in the CLI or Ctrl+T in the TUI.
Prior reasoning is now replayed by default on every provider that can read it, across resume, branches and compaction.
When a provider has a bad hour, Hermes waits out short 503, 529 and timeout errors in visible, bounded cycles instead of ending your turn.
🔥 Want the exact Hermes Agent setup I run inside my Agent OS?
Inside the AI Profit Boardroom you get the installable Agent OS, step-by-step video tutorials, four coaching calls a week and 3,400+ members building real automations.
How Hermes Agent v0.22 fits an Agent OS workflow
My own Agent OS runs Hermes as one of its agents, and I keep one Hermes profile per model.
That is exactly the setup the new single gateway is built for.
| Agent OS job | The v0.22 feature that helps | Why it matters |
|---|---|---|
| Running a profile per model | One backend and one gateway serve every profile. | New profiles go live without a restart, and each keeps its own keys, cron jobs and sessions. |
| Agents that update your tools | Connectors for hosted apps such as Linear and Notion. | The agent can act in the same turn after one browser sign-in. |
| Private or zero-cost jobs | One-click local models on llama.cpp. | Hermes picks a model that fits your GPU and RAM, with no key. |
| Long research and build runs | Reasoning replay and bounded outage waits. | A bad provider hour no longer ends the turn. |
| Overnight scheduled jobs | Self-healing cron schedules. | A run that never reached the model retries after 5, 15 and 30 minutes. |
| Delegating to subagents | Live subagent view with Steer and Stop. | You can correct a child before it wastes a run. |
| Logged-in browser work | Password-manager sign-in and Bot Screen. | The model never sees the password, and you can take the screen for a 2FA step. |
The way I would roll it out is one change at a time.
Update first, then connect one app, then add one local model, and only then switch on Bot Screen.
If something breaks, you will know which change caused it.
The models you can pick in v0.22
Nous Research calls this the Claude 5.5 and GPT-6 model wave.
| Family | Models the release notes list |
|---|---|
| Claude 5.5 wave | Claude Fable 5.1, Opus 5.5, Sonnet 5.5 and Haiku 5.5. |
| GPT-6 wave | GPT-6 Astra, Sol and Luna, plus GPT-6.1 Sol. |
| Gemini | Gemini 4 Argon, and Gemini 3.8 Flash in the model catalog. |
| Local catalog | Qwen3.8 27B, Qwen3.8 Flash Next, Qwen3.6 35B-A3B and DeepSeek V4 Flash, plus any Hugging Face GGUF. |
| Also added | DeepSeek V4.1 Flash, Qwen3.8-Max-0902, Grok 4.7, GLM-5.3-FlashX, Muse Spark 1.3 and Solar Pro 4 and Mini 4. |
They are pickable on Nous Portal, OpenRouter and native routes where the provider offers them.
The model picker now shows a chip such as "17% left · resets 3:00 PM" before a subscription runs dry, and you can set reasoning effort with /model X --reasoning high.
v0.22.1 adds a curated model list at the top of the CLI picker, which makes the long list easier to scan.
For my picks by job, read the best Hermes Agent model.
The smaller changes worth knowing
All user-facing text now runs through one translation catalog with full coverage in 16 bundled languages.
About 19 rarely used tools now load only when the agent searches for them, which roughly halves the desktop tool schema on every call.
Ctrl+C now clears your draft instead of killing the turn, and /queue is editable.
Unpinned cron jobs now use your main model at the moment they fire.
The desktop app has a Simple mode that hides the developer panes.
What the v0.22.1 patch changed
Nous Research released v0.22.1 on 14 October 2026 as a patch with a handful of fixes for the Windows Store, plugin updates and more.
It adds the curated model list at the top of the model picker.
It also adds catalog plugins, including Korean and Brazilian Portuguese language packs for the desktop app.
Also On Our Network
FAQ
What is Hermes Agent v0.22?
Hermes Agent v0.22 is a major release of the free, open-source Hermes Agent from Nous Research, called The Threshold Release. The release notes date it 13 October 2026, and it adds a setup chat, connectors, one-click local models, stable releases, signed desktop packages and a 600-plugin catalog.
How do I update to Hermes Agent v0.22?
The desktop app updates itself, the Microsoft Store version updates through the Store, a command-line install updates with hermes update, and Docker users pull nousresearch/hermes-agent:stable.
Does Hermes Agent v0.22 need an API key?
Not for the first-launch setup on a fresh desktop install, which runs on the free nous/welcome trial model once per operating-system user. For real work you connect a provider or install a local model, which needs no key.
Which models can I use in Hermes Agent v0.22?
The release notes list Claude Fable 5.1, Opus 5.5, Sonnet 5.5 and Haiku 5.5, GPT-6 Astra, Sol and Luna, GPT-6.1 Sol and Gemini 4 Argon, plus local models such as Qwen3.8 27B and DeepSeek V4 Flash.
What changed in Hermes Agent v0.22.1?
v0.22.1 is a patch from 14 October 2026 with Windows Store fixes, a curated model list at the top of the model picker and new catalog plugins, including Korean and Brazilian Portuguese language packs.
Will updating delete my old Hermes chats?
Ended sessions that have been inactive for 90 days are now deleted at startup by default. Pinned and open sessions are never touched, and setting sessions.auto_prune: false keeps everything.
Final word on Hermes Agent v0.22
Update, read the upgrade notes, then add connectors, a local model and the plugins you need one at a time.
If you run an SEO business and want to talk through where agents fit, you can book a free SEO strategy session with my team.
If you want the whole Agent OS with the Hermes tab, the profiles and the training already set up, it lives inside the AI Profit Boardroom, with 3,400+ members and four coaching calls a week.
That is the fastest way to get real work out of Hermes Agent v0.22.
About Julian
I'm Julian Goldie, an AI entrepreneur, SEO expert and the founder of the AI Profit Boardroom, which has 3,400+ members.
I help business owners scale with AI agents, automation and SEO.
My YouTube channel has 400,000+ subscribers, and I run Goldie Agency, a seven-figure SEO agency.
Related guides on agentos.guide
→ The Best Hermes Agent Plugins
Get the whole Agent OS
The installable system behind every guide on this site — dashboard, agents, memory and pipelines — updated as new tools land, with 3,400+ members and four coaching calls a week.
Join the AI Profit Boardroom →📺 Video notes + links to the tools 👉
🎥 Learn how I make these videos 👉