Skip to content
Vestra
← All comparisonsReviewed

Vestra vs Hermes Agent

Vestra gives every person a chief they name, who hires and runs a whole team of AI teammates that hold whole jobs and work together in one office, each on the engine that suits its job, with nothing to install. Hermes Agent is a free, open-source agent from Nous Research that you install, configure and host yourself (or on Hermes Cloud), with the model you pick and Bot Mode for running several bots of your own.

Hermes Agent

  • One agent you install and run. Free and MIT-licensed, on your own computer, server or Hermes Cloud, set up with the model and tools you pick.
  • Bots for one person. Bot Mode turns your profiles into named bots with their own model, memory and skills, which can message each other in group chats.
  • Free, plus model costs. Use local models, any provider, or a Nous Portal plan from $20 a month with one Hermes Cloud agent.
  • Lives in your chat apps. More than 20 messaging platforms, including Telegram, Slack, WhatsApp, Signal, Microsoft Teams and iMessage, each set up by you.

Vestra

  • A chief that runs a whole team. Name your chief, hire teammates in a sentence, give each a whole job, and watch them all at work in an office you can walk around.
  • From $19, for a team. Plans start at $19 a month for a chief and up to 3 teammates, hosting included. Start with 1,000 free credits.
  • Calls, email and the apps you use. Your chief has its own number and email address, and works in Slack, Telegram and the app. Connect 2,000+ apps in a click.
  • Limits, approvals and receipts. Cap spending per task and per day, approve anything big, and see what every piece of work cost.
The Vestra office from above: team desks on coloured rugs, a café, two glass meeting rooms, a row of phone booths, a lounge and a library, with the front door at the bottom.

Can you see what your whole team is doing?

Yes. Your Vestra office is a map you can walk around. Every person, chief and teammate has a desk, and a Right now feed says in one line what each one is doing, live. Walk over or tap any desk to see its work, watch its computer and take over.[1]

Vestra's four plans side by side: Lite at $19 a month for up to 3 teammates and 2,000 credits, Pro at $97 for 20 teammates and 10,000 credits, Max at $197 for 50 teammates and 20,000 credits, and Ultra at $497 for 150 teammates and 50,000 credits.

What does a whole team cost, with nothing to host?

Vestra starts at $19 a month for a chief and up to 3 teammates. Lite includes 2,000 credits every month, every new office starts with 1,000 free credits, and unused credits roll over. Pro, Max and Ultra add up to 150 teammates. Hosting is included on every plan.[2]

Today's receipts in Vestra showing what four teammates each spent on their jobs, 140 credits in total, next to a daily cap of 400 credits, a per task cap of 80 credits, and a first email to a new lead waiting for your yes.

Can it call, email and stay inside your limits?

Yes. Vestra’s chief has its own phone number and email address. Phone calls start on Pro.[3] You cap spending per task and per day, money and first contact wait for your yes, and every piece of work has a receipt.[4]

Hand your first job to your chief.

Start with one real job you want off your plate. See how your chief and its teammates handle it, then build from there.

  1. Sign up and start with 1,000 free credits.
  2. Name your chief and tell it about your work.
  3. Hire your first teammate in a sentence and connect the tools it needs.
  4. Review the finished work, and set the limits and approvals you want.

Vestra vs Hermes Agent, in detail

The rows follow the four reasons above: a chief that runs a whole team; from $19 for a team; calls, email and the apps you use; and limits, approvals and receipts. After those come the areas both products cover, and how each one approaches them. Hermes Agent is a free, open-source agent that one person installs, configures and runs, with bots of their own. Vestra gives each person a chief and a whole team of agents on different engines, working together as teammates in one office, with nothing to host.

How the products compare: Vestra and Hermes Agent
What mattersVestraHermes Agent
Who runs the workYou name your chief and tell it what you need in a sentence. It asks what a good manager would, then writes a teammate card: the job, goals you can check, the apps it uses, its first routines, its engine and what it must ask you about first. You read it, change anything and say yes. From then on your chief hands out the work, makes sure every result is checked before it reaches you, and brings you only real decisions. Nothing to install or configure: you get a chief and up to 150 teammates, each holding a whole job, working together in one office.[5][6]You do. Hermes Agent is one agent you install and set up yourself: you pick its model, write its persona in a SOUL.md file and choose its tools. Bot Mode in Hermes Desktop turns your profiles into a roster of named bots, each with its own role, model, memory and skills, and bots can message each other and share group chats where a lead bot or an automatic router picks who answers. Every bot is one of your own profiles on your own machine or Hermes Cloud, and its Kanban docs leave auto-assignment, budgets and org-chart views for you to build.[7][8][9]
See your whole teamYour office is a map you can walk around, like a game. Every person, chief and teammate has a desk, and a Right now feed says in one line what each one is doing, live. An orange tag shows who is waiting on whom, and each teammate’s tag carries the logo of the engine it runs on. Tap any desk to see its work, watch its computer live or take over. Nothing moves on the map unless real work is happening.[1]Hermes Desktop shows a Bots roster with each bot’s avatar, latest message and an Active now filter, and a Kanban dashboard shows tasks moving between your profiles. On a Linux server or Hermes Cloud, Bot Screen streams a bot’s own desktop so you can watch it and take over for a login, then hand back. The docs note that bot screens are work surfaces, not security boundaries: bots on one host share its user account, files and network.[7][10]
PlansLite $19 a month for up to 3 teammates and 2,000 credits. Pro $97 for 20 teammates, 10,000 credits and phone calls. Max $197 for 50 teammates and 20,000 credits. Ultra $497 for 150 teammates and 50,000 credits. Billed yearly, they are $17, $87, $177 and $447 a month. One credit is one cent, the same for model time, the computer, searches and calls, and every plan includes your chief, the whole office and the hosting. Unused credits roll over. Enterprise from $4,000 a month.[2]Hermes Agent is free and open source under the MIT license, and local models cost nothing to run. Otherwise you pay a model provider, or use a Nous Portal plan: Free ($0, free models only), Plus $20 a month, Super $100 and Ultra $200, each paid plan with hosted tools (search, images, speech, a cloud browser) and one always-on Hermes Cloud agent. Bigger cloud sizes and extra agents use credits, billed daily on usage. Hermes Business runs on one central team balance with per-member spend caps; no price is listed.[11][12][13]
Getting startedSign up directly, on any plan, with nothing to install or host. Every new office starts with 1,000 free credits, enough to hand over real work before you pick a plan. Your chief suggests your first teammates, you connect a few tools, and work starts the same day. Your team’s memory fills from your tools’ history on day one, so it already knows your customers. In India, plans are billed in rupees with GST included.[2][14][15]Download Hermes Desktop for macOS or Windows, or run the install script on macOS, Linux or Windows, then connect a model provider; hermes setup --portal does it with one Nous Portal sign-in. For an always-on agent, deploy to Hermes Cloud from Nous Portal, or host it on your own server. hermes migrate previews and imports an existing OpenClaw, Claude Code, Codex CLI, Cursor, Grok Bot or OpenBot setup.[11][16][8]
Where you use itTalk to your chief in the Vestra app on the web and the desktop app for Mac, in Slack (by direct message or a mention in any channel), in Telegram, or by email: forward anything, or copy it in on a thread. It’s the same chief everywhere, so a conversation started in Slack can finish in the app. Decisions reach you the same way wherever you are, and replying “1” takes the suggested option. Nothing to host: there are no tokens or webhooks to wire up.[3][5]More than 20 messaging platforms, including Telegram, Discord, Slack, WhatsApp, Signal, Microsoft Teams, Matrix, LINE, SMS, email and iMessage, plus the CLI, Hermes Desktop and a web app served from your own gateway. One gateway per machine owns every conversation, so a chat continues across surfaces. Each platform is set up by you, with your own bot tokens and allowlists. A mobile app is in beta.[17][8]
Calls and emailYour chief has its own phone number and email address from day one, so it calls and writes from its own line and inbox, not yours, with no accounts or webhooks to set up. It writes first, replies and follows up until there’s an answer. Before a call it reads what the team knows about the person, and afterwards the notes and next steps go where they belong. The first call or email to anyone new waits for your yes. Phone calls start on Pro.[3][2][18]Email works through an account you create for the agent and connect over IMAP and SMTP; the docs recommend a dedicated address, not your personal one. Texts and calls run through your own Twilio account: the SMS gateway needs a Twilio number and a public webhook, and an optional telephony skill provisions numbers and makes AI outbound calls.[19][20][21]
App connectionsMore than 2,000 apps connect in a click, including Gmail, Slack, HubSpot, Notion, Shopify and Stripe. Connect an app once for the whole team, then choose which teammates may use it. Logins stay sealed: agents use them without ever seeing the password, and you can take access back in one tap. Where there’s no connector, a teammate uses the website on its own computer.[18][4]Any MCP server, local or remote, with a curated one-click catalog, plus bundled and optional skills and more than 1,000 plugins in the Hermes Marketplace. With a paid Nous Portal plan, a Tool Gateway adds web search, image generation, speech and a cloud browser without separate keys. Website logins are kept in an encrypted vault and filled into the page without the model seeing them, and 1Password or Bitwarden are picked up if installed.[22][23][16]
Model and engine choiceMore than a model: pick the engine that does each teammate’s work, Vestra’s own, Claude Code or Codex. Your chief suggests one and says why, or picks job by job. Mix them in one team, so code written on Claude Code can be reviewed on Codex. Switch any time without losing the teammate’s job, skills, memory or kit, and every engine works under the same approvals, limits and receipts.[24][3]Hundreds of models: 200+ through Nous Portal, any OpenAI-compatible API, a ChatGPT or SuperGrok sign-in, or local models through Ollama and similar. Each bot can be pinned to its own model. Bundled skills hand coding to the Claude Code and Codex command-line tools, and an opt-in runtime runs OpenAI turns inside Codex’s own app-server. Hermes stays the agent on top; the other tools are called as helpers.[8][7][25]
Its own computerYour team works on a cloud computer with a real browser, files and a terminal, run and looked after by Vestra. It uses any website the way you do, including ones with no connector, like supplier portals, booking pages and government forms. Watch it live, take over with one click when a site needs you, and hand it back to carry on where you left off. It has 8 GB on Lite and Pro, 16 GB on Max and 32 GB on Ultra.[18][2]Hermes runs on your hardware or a Hermes Cloud container, with a terminal, files and five sandbox backends (local, Docker, SSH, Singularity, Modal). Computer use drives your own desktop in the background on macOS, Windows and Linux with any tool-capable model. Hermes Cloud runs 4 GB to 8 GB of RAM on the Portal plans, and you keep the machine running yourself when you self-host.[11][26][12]
ApprovalsSpending over your limit, contacting someone new and deleting anything always come to you, and no setting turns that off. Everything else you tune teammate by teammate and task by task: ask first, suggest, or act and tell you afterwards. New teammates start by asking, and get more freedom only when you move the setting. Each decision arrives in one line with the facts and a suggestion, so you can answer from your lock screen.[4][5]Approvals cover dangerous shell commands. The default smart mode lets a second model approve low-risk commands, deny clearly dangerous ones and ask you about the rest; manual asks every time, and off (or /yolo) skips the prompts entirely, except a hard blocklist. Scheduled jobs deny dangerous commands by default rather than wait. Who may answer an approval in a shared chat is a setting you choose.[27]
Spending limits and receiptsCap any single task and the whole day, up to $100 a day. Anything unusually big, over 50 credits by default, waits for your okay with the cost shown. When a cap is reached, work stops at a sensible point and says what’s left. Every receipt splits the cost into model time, actions and computer time, and shows which teammate spent it on which job. Unused credits roll over, and you can top up any time.[4][2]An MCP server can declare what each tool costs and a spend limit per chat; a call that would cross it asks first, even in YOLO mode. The running total lives in memory for one conversation, “not the vendor’s bill”, and resets when Hermes restarts. Hermes Business adds per-member spend caps on a team balance. The docs we read don’t describe a daily cap or an itemised receipt for each piece of work.[22][13]
MemoryA shared wiki of your people, customers, decisions, numbers and what went wrong last time, with a source for every line. It fills from your tools’ history on day one, then from every task, email and correction. Open any line’s source or edit any page, and the fix reaches every teammate’s next piece of work: correct one teammate once, and they all learn. Your private notebook stays yours. Export it all as Markdown that opens in Obsidian.[15][3]Each profile keeps two small files, MEMORY.md (2,200 characters) and USER.md (1,375 characters), which load into every session, plus searchable past sessions and skills it writes for itself. Memory is per profile by design: the docs say not to point two agents at one Hermes home, and to use an external memory provider if they need shared memory. Everything is stored on your machine in ~/.hermes.[28][29]
Schedules and proactive workEach teammate owns a goal and does three kinds of work: one-off asks, routines that run on a schedule (a market note every Monday at 9) or the moment something happens (a rival changing its prices), and the goal it holds for months. When it spots something worth doing, it brings you a proposal to approve, and on quiet days it says nothing. Your chief sends a morning brief at the time and place you choose.[6][5][6]Cron jobs in plain language or cron syntax, triggered by webhooks as well as time, with results delivered to the chat they came from. In Bot Mode each bot’s routines sit beside its chat. /goal keeps one session working until a judge model says the goal is met, and /heartbeat repeats one instruction in a conversation.[30][31][7]
SkillsReady-made skills for Sales, Bookkeeping, Content, Customer care and Research, added to any teammate in a click. Custom skills are built around your own playbooks, data and standards, usually during onboarding, and read like a short plain-language checklist you can edit when your business changes.[6]Skills are documents in the open agentskills.io format that Hermes loads when needed, and it writes new ones from work it has done. Bundled, optional and community skills install from the Skills Hub with a security scan. On Hermes Business, skills any member builds are published to a team library for everyone’s agent.[32][13]
ResearchHire a Research teammate in a sentence. Findings land in your memory wiki with a source for every line, so the next teammate starts from them, and the rest of your team turns them into the outreach, the deck or the price change.[6][15]Web search and extraction, browser automation and vision are built in, and delegate_task splits research across subagents with their own context that report back a summary.[11][33]

Frequently asked questions

Is Vestra a good Hermes Agent alternative?

A whole team, not one agent you run. Every person gets a chief that hires teammates in a sentence, gives each a whole job and checks their work before it reaches you.[5]

An office you can walk around. See every teammate at its desk and what it’s doing right now.[1]

An engine per teammate. Run each teammate on Vestra’s own engine, Claude Code or Codex, under one set of rules.[24] For how engines work and how to pick one per job, read What is an AI agent engine?[34].

From $19, nothing to host. Vestra plans start at $19 a month for a chief and up to 3 teammates, hosting included.[2] Hermes Agent is free; models and Hermes Cloud are paid, with Nous Portal plans from $20.[11]

Calls, email, limits and receipts. Your chief has its own number and email address, connects to 2,000+ apps, and works inside per-task and daily caps with a receipt for every piece of work.[4]

When does Hermes Agent make more sense?

If you want an agent you own end to end, Hermes Agent is hard to beat: it’s free and MIT-licensed, runs on your own hardware or fully offline with local models, works with hundreds of models, and answers on more than 20 messaging platforms, WhatsApp, Signal and iMessage included. Developers who like to configure everything, from the persona file to the sandbox, get a deep toolkit and a large plugin marketplace. If one capable agent for yourself is what you need, and you’re happy to set it up and look after it, Hermes may be all you need.[8]

What is Hermes Agent?

Hermes Agent is a free, open-source AI agent from Nous Research, under the MIT license. You install it on your computer or server, or deploy it to Hermes Cloud, connect the model you pick and talk to it from the desktop app, the terminal or your chat apps. It remembers your projects, writes its own skills and runs scheduled jobs, and Bot Mode lets one person run several named bots.[11]

How much do Vestra and Hermes Agent cost?

Vestra starts at $19 a month. Lite is $19 for up to 3 teammates, Pro $97 for 20, Max $197 for 50 and Ultra $497 for 150, with 2,000 to 50,000 credits a month. One credit is one cent, unused credits roll over, and every new office gets 1,000 free credits.[2]

Hermes Agent itself is free. You pay your model provider, or use a Nous Portal plan: Free, Plus $20 a month, Super $100 or Ultra $200, each paid plan with one Hermes Cloud agent. Extra or bigger cloud agents use credits, billed daily on usage.[11]

Doesn’t Bot Mode already give me a team?

It gives one person a roster of bots. Each bot is one of your own Hermes profiles, with its own model, memory and skills, and you set each one up, keep its machine running and decide who answers in a group chat. Memory isn’t shared between them unless you add an external provider. In Vestra, every person has a chief that hires and runs a whole team: teammates that each hold a job on the engine you pick, share one memory with a source for every line, and work in one office with the same limits, approvals and receipts.[7]

Does Vestra ask before it spends or sends?

Yes. Spending over your limit, contacting someone new and deleting anything always come to you, whatever your settings. You cap spending per task and per day, up to $100 a day, and every piece of work has a receipt. Hermes Agent asks before dangerous shell commands in its default mode, can be switched to never ask with YOLO mode, and lets an MCP server declare a spend limit per chat; the docs we read don’t describe a daily cap.[4]

Can Vestra run work on a schedule?

Yes. Routines run on a schedule, or the moment something happens in a connected tool. Your chief also sends you a morning brief, and teammates bring you proposals when they spot something worth doing.[6]

Can I see and fix what Vestra remembers?

Yes. Vestra keeps a wiki of what your whole team knows, with a source for every line, and you can open, check, edit and export it. Hermes Agent keeps two short memory files per profile on your machine, which you can read and edit, plus the skills it writes.[15]

Can I move my Hermes Agent setup to Vestra?

Write down your cron jobs, your bots and the skills you use most. Then describe each recurring job to your chief in a sentence, so each one gets its own teammate, reconnect the apps it needs, and start with one familiar task so you can check the result against how you already work. Your memory wiki fills from your tools’ history on day one.[6]

How do I get started with Vestra?

Sign up, name your chief and tell it about your work. It suggests your first teammates, you connect a few tools, and work starts the same day. Every new office starts with 1,000 free credits, and plans start at $19 a month.[14]

A whole team, not one agent to run, from $19.

Hermes Agent gives you one free, open-source agent to install, configure and run yourself. Vestra gives every person a chief that hires and runs a whole team, each teammate on its own engine, with calls, email, limits and receipts and nothing to host, from $19 a month.

Get started with Vestra