Plans and AI credit

You pay for two separate things:

A monthly plan that sets your capacity: how many workspaces, agents, and reports your billing account can hold, and how many report runs it can execute per month.
Prepaid AI credit that covers what the AI actually does. Everything that uses AI draws from it: chat, agent runs, and the builders that create and edit your reports and agents. The plan covers infrastructure, hosting, scheduling, and delivery.

The split keeps both sides predictable. The plan price never changes with usage, and AI spending can never exceed what you have deliberately topped up.

Plans

	Starter	Pro	Business
Price	49 € / month	149 € / month	299 € / month
Workspaces	1	3	10
Agents	5	20	50
Reports	5	20	50
Report runs / month	1,500	6,000	15,000

All allowances are pooled across the billing account's workspaces. Draft agents do not occupy a slot; a report occupies a slot once it has a saved version. If you need more than Business, contact us for custom terms.

Changing plans, payment, invoices

Plan switches, the payment method, and all invoices are handled through Manage billing on the billing account page, which opens the Stripe customer portal. Cancelling takes effect at the end of the paid period; you keep full access until then and can resume before it ends.

If the subscription lapses

When a billing account has no active subscription, a banner appears in the app and everything that executes AI work pauses: chat, report runs, and agent runs. Nothing is deleted, and viewing, configuration, and billing remain fully available. Subscribing again lifts the pause; schedules simply resume.

AI credit

AI credit is a prepaid wallet in euros, held at the billing account.

What draws credit: everything that uses AI. That means chat conversations, agent runs, the builders' work when you create or edit reports and agents with AI, and automatic chat titles.

What does not: scheduled report runs and deliveries. They run the saved report without an AI model, so they are covered by the plan.

You receive 25 € welcome credit with your first subscription.
Top up from the billing account page: 20, 50, or 100 €, or a custom amount (minimum 20 €). Credit never expires.
Auto-reload is optional: when the balance falls below a threshold you set, the card saved with your subscription is charged a fixed amount. If a charge fails repeatedly, auto-reload switches off and the owner is emailed.
When the balance reaches zero, chat, agent runs, and the builders pause until you top up. Scheduled reports keep running, since they do not use credit. Nothing is lost.

Every euro is attributable: usage is shown per feature, per agent, and per workspace under Settings → Cost Management and on the billing account's Usage tab. To cap spending ahead of time, see Limits & cost controls.

AI tiers

Not all AI models are alike. Some are fast and inexpensive, and they handle everyday questions well. Others are smarter and better at demanding work, such as building a report or setting up an agent, but they take longer and cost more. Discontinue groups its models into tiers, from quick and inexpensive to smart and more expensive. Which tiers are offered, and the models behind them, change as better models become available; the current tiers and their prices are listed under Model pricing.

Which tier is used

Each feature chooses its tier on its own:

Chat: when you start a new chat, you can pick a tier with the gauge icon in the message box. The chat keeps that tier for all of its replies. If you don't pick one, the chat uses the workspace default for chat.
Agent runs: each agent can have its own tier, set on the agent. Without one, its runs use the workspace default for agents.
Report builder and agent creator: always use their own workspace default. This also applies when you ask for a report or an agent in a chat: the tier picked for the chat covers the chat's replies, not the report builder or agent creator the chat hands the work to.
Chat titles: the automatic titles use a small tier chosen by the platform and cannot be changed.

The tier choice in the message box and on the agent only appears when more than one tier is offered for your workspace region.

Changing the defaults

The workspace defaults are under Settings → Cost Management → Models, one each for chat, agents, the report builder and the agent creator. Every member can see them; workspace admins can change them. A feature without its own workspace default follows the platform default, shown there as "Platform default".

A quicker tier keeps costs down for everyday work. A smarter tier pays off where the result matters most, for example in the report builder and agent creator, whose reports and agents you keep using long after they are built.

Model pricing

All AI work draws credit from its measured token usage. Each rate below is in EUR per 1 million tokens and covers one of the four token classes that can appear in a run: input, output, cache read, and cache write.

Choose EU or Global to see the AI tiers currently offered for that workspace region. A tier can be offered in both regions with its own rates in each; the table keeps each tier's published name and rates. This reference is available without signing in.

Loading the current public model prices…

Every feature is charged at its tier's rates for your region; AI tiers above explains which tier each feature uses.

The rates are net prices excluding applicable VAT, just like the monthly plan prices above. Your usage views show the resulting euro charges by feature, agent, and workspace.