Models are rented.
Your Brain is owned.

Your module price is a platform fee for what connects the two: a Brain that keeps your data and know-how, and agents you approve. Normal model usage is included. Beyond it, tokens are billed at the provider’s list price, with no markup.

MODELS · RENTED · AT COSTClaudeLIST PRICE · 0%GPTGeminiOpen-weightSAFEGUARDS · ACCESS · APPROVALS · LOGYOUR BRAINdata + know-how + decisionsQUOTESREPLIESREORDERSTHE PLATFORM FEEYOUR DATA · OWNEDCRMOrdersDocsEmailDecisions

Rented, and owned

Rented: the models
  • Billed at the provider’s list price, never marked up
  • Swapped when a better or cheaper one arrives
  • Cheaper every generation
  • Never trained on your data
Owned: your Brain
  • Your data: CRM, orders, documents, email
  • Your know-how and every decision you approve
  • More useful every month
  • Exports any time

What the platform fee pays for

Each task goes to the model that does it best for the price. Changing a model is a setting; your data doesn’t move.

Sort a replyFAST MODEL$Draft a proposalSTRONGEST MODEL$$$Tag 5,000 productsOPEN MODEL¢MODEL PICKERONE SETTING PER TASK · YOUR DATA NEVER MOVES
Your Brain
One place for your data and know-how: customers, products, prices, documents, and every decision you approve. It’s yours, and it exports any time.
Picking the model
Claude by default; GPT, Gemini, Grok, Mistral, open models or Jev where they do better, cheaper, or in your region.
Agents you approve
Each agent gets only the access its job needs. Anything that spends money, changes a price or goes public waits for a person, and every action is logged.
The people behind it
Setup for your business, tests before every model change, and updates when a provider changes its prices or retires a model.

Tokens: included, then at cost

A token is the unit model providers charge for reading and writing text.

MODEL USAGE THIS MONTHINCLUDED IN YOUR MODULESBEYOND: LIST PRICE · 0%ONE LINE PER MODEL ON YOUR INVOICEMODELTOKENSPUBLISHED PRICEADDED: 0%
Included
Normal model usage for each module is part of its monthly price, at the volumes listed on the module’s page.
Beyond that, at cost
Anything past the included volume is billed at the provider’s list price, with nothing added, one line per model.
Your own account
Connect your own provider account (Anthropic, OpenAI, Google and others) and pay them directly. Your platform fee doesn’t change.
Over 200 people
Model usage is billed at cost from day one, next to a scoped engagement for your organization.

Why we don’t mark up tokens

Prices fall fast
Model costs drop from one generation to the next. Without a markup, you get those drops the day they land.
The right incentive
A margin on tokens would pay us to use bigger models. Our fee is flat, so our interest is the smallest model that does the job well.
You can check it
Every line shows the model, the tokens and the provider’s published price. You can redo the math.
Same rule as ad spend
Ad budgets are paid to the platforms directly and never marked up. Tokens follow the same rule.

Value that compounds, whatever the model

Models change every few months. What your business teaches its agents stays.

YOUR BRAINMONTH 1 → 12Today’s modelBetter modelTESTED SWITCH
Every approval teaches
Approved drafts, your edits and what worked go back into your Brain. Next month starts from there.
The knowledge is yours, not the model’s
Your Brain lives outside the model. When a better model arrives, it starts from everything your business already taught the last one.
Every switch is tested
Before a switch, the old and the new model answer the same cases from your business, and the results are compared. We do the same at home: the way Humonix sorts replies to its own outreach was chosen that way.

Questions

What is the platform fee?

Your plan’s monthly price: AI visibility at $95, Hue Pro at $350, One module at $350 or Head of Digital + Taylor at $6,000. It pays for your Brain, picking the model, the agents you approve and the people who set them up.

Which model usage is included?

What each module needs at its listed volumes: for example 20 posts a month for Social or 4 articles for Content. Each module’s volumes are on its page.

What does “at cost” mean?

The provider’s published list price for the model and tokens used, with nothing added. Applicable taxes apply as on the rest of the invoice.

Can I use my own provider account?

Yes. You then pay the provider directly, and your platform fee doesn’t change.

What if a provider changes its prices?

Included usage stays included at the price you signed. Usage billed at cost follows the provider’s list price, up or down: price cuts reach you the same day.

Do you pick models to save yourselves money?

No. We choose on quality first, tested on your own cases, then on price. Since our fee is flat and we don’t mark up tokens, a pricier or a smaller model changes nothing in what we earn.

Is my data used to train models?

No. Your data is never used to train shared models, and you can export your Brain any time.

See what your usage would cost

On a call we estimate model usage for the modules you’re looking at, and what would be included.

Book a call