"I am not technical. I do not know how to install it."
The install
Hermes tutorials assume you can rent a VPS, open a terminal and keep a server patched. You should not have to. We install it, update it and watch it for you.
New plan · Tessera Founder
Your private AI assistant, installed and working. $55 a month.
For founders and one-or-two-person companies. We install and host your Hermes assistant, connect it to your Telegram, WhatsApp and other platforms, and teach you how to get results alongside a community of founders like you. Flat fee, no surprises.
What we hear every week
You want an AI assistant. Three things stop you.
"Everyone says you should run it in a private environment."
The security
They are right. An assistant with access to your messages should not run just anywhere. Your Hermes lives in an isolated instance, yours alone, on servers in Europe with an encrypted disk. What the tutorials recommend, done by us.
"I am afraid of burning tokens like crazy."
The bill
A Hermes doing real work consumes fast: connected to a commercial AI, the bill changes every month and can jump to hundreds of dollars. With Tessera that is impossible: a flat $55, no meter. Your assistant works as much as it needs to.
Why this plan exists
We built it for ourselves first.
My co-founder and I have been running personal assistants (Hermes and OpenClaw) connected to our Telegram for a long time. They read our email, prepare reports, research markets, schedule tasks. They work so much that on a commercial AI each of us would pay thousands of dollars a week. That is why we built Tessera: open models, servers in Europe, a fixed cost.
One day we showed our assistants to two founder friends. Not technical. They saw what the assistants did for us and wanted the same, but they were not going to rent a VPS or fight a terminal. So we set it up for them: their private Hermes, connected to their WhatsApp, running on Tessera. And we kept passing them our use cases and our skills.
Today they are the first two members of the Founder plan. This page opens those same seats to more founders.
Jose Medina, founder of Tessera
How it works
Three steps. It answers within your first week.
-
Step 1: Apply for your seat
Tell us what you work on and what you want to delegate. We confirm the plan fits you, and if it does not, we say so.
-
Step 2: We set up your Hermes
A private instance on European servers. Telegram is connected on day one; WhatsApp through the official Meta API during onboarding. You leave the 1:1 kickoff call with your assistant answering on your phone.
-
Step 3: Learn with the community
A founders Discord in English and Spanish, a monthly live training session and a weekly audio podcast. New skills land on your assistant every month, many born from other members’ real workflows.
What is included
Everything you need to work with AI. One fee.
Five pieces you would otherwise pay five different vendors for. Under each one, its real market price.
Your private, managed Hermes
Hermes is the reference open-source personal agent, OpenClaw style. You get an isolated instance that is yours alone: we install it, update it and watch it.
Bought separately
Managed hosting elsewhere: $20 to $60/mo, plus your time
On your Telegram, WhatsApp and more
Text it or send a voice note: it transcribes, replies in text or audio, and takes scheduled tasks. Telegram through the official API from day one; WhatsApp through the official Meta API during onboarding; plus other platforms like Discord or Slack.
Bought separately
Commercial assistants like Martin: about $21/mo, and not yours
Flat-rate AI, no meter
Your fee includes all five Tessera open models: conversation with reasoning, speech to text, text to speech and semantic search. No token meter: your assistant and you can use it around the clock. The only limit is pace, 20 requests per minute.
Bought separately
AI chat subscriptions: $20 to $100/mo, and not your infrastructure
Founders community
A Discord with separate English and Spanish spaces. Share your workflows, see everyone else’s, and the best ones become new skills for every member.
Bought separately
Paid founder communities: $30 to $50/mo
Monthly training and weekly podcast
One live session per month on a specific topic, recording available, plus a weekly audio podcast to stay current without reading twenty newsletters.
Bought separately
Ongoing AI training: $50 to $100/mo
Founding bonus
1:1 kickoff call with the founder
A video call with you to leave your Hermes connected and adapted to your business. You end the call texting your assistant from your phone. Included with founding seats.
The AI, included
It is not just Hermes. It is your flat-rate AI.
Your fee gives you direct access to all five Tessera open models, the same ones our business customers use. No token meter; the only limit is pace, 20 requests per minute, around the clock.
Converses and reasons
The brain of your Hermes. It writes, summarizes, analyzes and switches to deep reasoning when the problem calls for it. No monthly usage cap.
Understands your voice
Transcribes your voice notes and meetings in English, Spanish, Portuguese or Catalan. Send it an audio on your way to a meeting and have it done when you arrive.
Talks back
Your assistant replies in audio when you are driving. It is the same model we use to produce the community’s weekly podcast.
Finds what you cannot remember
Semantic search over everything you tell it and send it: your Hermes finds "that document Ana sent me" even if you cannot recall its name.
And the same key works outside Hermes: connect it to opencode for coding, to your scripts or to any tool compatible with the OpenAI API. One payment for all the AI you use.
The honest math
What assembling this yourself would cost.
Approximate public prices as of July 2026. You can verify every line.
The honest version: you do not need every piece at once. But with two or three of them you already pay more than the whole plan, and none of them is yours.
Pricing
One fee. No surprises.
$55/mo
or 55 € if you are in Europe · Less than $2 a day.
- No lock-in: cancel anytime
- Money-back guarantee on your first month
- Founding price locked forever
- Optional annual billing: $47/mo ($564 a year, save $96)
Founding price. We are opening 25 founding seats in this first round. Join now and keep $55/mo for as long as you stay. Once they fill, the price for new members goes up and we will publish it here. No fake countdowns.
Apply for a founding seat We reply within 24 business hours.
Honest fit
Who this is for. And who it is not for.
It fits you if
- You are a founder or a one-or-two-person team and you want AI to actually take work off your plate, not to collect another tool.
- You live in Telegram or WhatsApp and want your assistant there, not in yet another tab.
- You want a fixed cost: knowing what you pay this month and the next.
- You care where your data lives and prefer a private instance in Europe over pasting customer information into a public chatbot.
- You want to learn from other founders who already use this daily.
It does not fit you if
- You need to serve AI to your product’s users at scale. That is our Async, Lite or Pro plans, and we help you choose.
- You need specific closed models (GPT, Claude, Gemini). We only serve open models.
- You already run your Hermes on your own VPS and enjoy maintaining it. You win there: keep going.
- You expect results with zero time invested. The assistant learns from you; the first week takes a few hours.
If your case is in the right-hand column, we tell you in the first conversation.
Frequently asked questions
What founders ask before joining.
Do I need to know how to code?
No. We handle everything technical: installation, WhatsApp and Telegram connection, updates and backups. You write messages, like to any other contact.
What exactly is Hermes?
An open-source personal AI assistant (MIT license), in the style of OpenClaw. It reads and answers messages, transcribes voice notes, researches the web, remembers what you tell it and runs scheduled tasks. We manage a private instance just for you.
Is this an OpenClaw alternative?
Yes. Hermes is an open-source personal agent in the style of OpenClaw, and we run it managed with the AI included. Typical OpenClaw hosting gives you the server but expects you to bring your own AI key and pay per token; here the model is included at a flat $55, with no meter.
What does it really cost to run an OpenClaw or Hermes agent?
The server is cheap: from $0 on a laptop or free tier, roughly $2 to $24 a month on a VPS or managed host. The cost that jumps is the AI: with your own key, an active agent can add tens to hundreds of dollars a month in tokens. With Tessera the AI is included in the flat $55.
Where does my data live?
In your isolated instance, on servers in Europe, with an encrypted disk. Your conversations are not shared with other customers or used to train models. We sign a DPA if you need one.
What does the 20 requests per minute limit mean?
Your assistant can make up to 20 AI requests per minute, sustained. For the real usage of a one-or-two-person company that is plenty. If your product grows and needs more, you move to a higher plan and keep everything.
Can I end up paying more than $55?
No. There is no token meter and no overage. The bill not moving is the reason Tessera exists.
Telegram or WhatsApp?
Both, with Telegram first. Telegram is connected on day one through the official API and has no per-message charge. WhatsApp connects through the official Meta API during onboarding, never the unofficial bridges that can cost you your number. Note that from October 2026 Meta charges per message for service replies on WhatsApp, so we recommend Telegram as the main channel. We also support other platforms like Discord and Slack.
Can I use the AI outside Hermes?
Yes. The subscription is an OpenAI-compatible API key, and Hermes is only one of its uses. Connect it to opencode for coding, to your scripts, to a spreadsheet or to any tool that accepts an OpenAI-compatible provider. All within the same fee.
What language is the community in?
There are separate English and Spanish spaces. The monthly training and the podcast are published in both languages.
How do I cancel?
With an email. No lock-in, no penalty. And if your first month does not convince you, we refund it. If you pay annually and cancel, we refund the full unused months.
Which AI models are included?
The full Tessera catalog, the same one as the business plans: Qwen 3.6 for conversation and reasoning (thinking mode with no monthly cap), Whisper for voice transcription, Kokoro for audio replies, and embeddings with a reranker for semantic search. All open models, no token meter. When we publish new models, you get them at no extra cost.
Your assistant can be answering this week.
$55 a month, flat. Installed by us, on your WhatsApp and your Telegram, with a community of founders behind it. First-month guarantee.