Team & Billing

Manage your organization, invite team members, assign roles, and handle plans, credits, and billing on botts.ai.

botts.ai is built for teams. From the dashboard you can invite colleagues, assign roles with different permission levels, choose where your data is processed, and manage your plan and credits — all under My Organization in the sidebar.

The My Organization area groups four pages:

• General — organization profile, billing identity, logo, feature toggles, internal chat visibility, and server location.
• Members — your members and pending invites.
• Plans — plan selection, credits, packages, auto-recharge, and invoices.
• Usage — credit balance and usage analytics.

Organizations

Every botts.ai account belongs to an Organization — the shared workspace where your team collaborates on agents, knowledge bases, deployments, and conversations. When you sign up, an organization is created for you.

Each organization has an ID (its slug, shown read-only) and a set of profile fields you can edit on the General page. Only owners and admins can open the General page; for everyone else it is hidden.

General settings

On My Organization → General, owners and admins can edit:

• Organization Name — used across the dashboard. Names containing <, >, or http(s):// URLs are rejected to prevent phishing (bare domains are allowed).
• Logo — upload an image for your organization.
• Billing identity — billing email, name, country, address, postal code, and city. Swiss addresses offer autocomplete, and changes sync to your Stripe customer record.
• Notification Email — the address that receives organization notifications.
• Features — toggle Regular Chat and Endless Chat on or off.
• Internal Chat Visibility — the organization default for new internal AI chats: private or full. This applies forward-only; existing chats keep the visibility they were created with.
• Server Location — your data-residency tier (see Data residency).

Deleting an organization is owner-only.

Team members

Roles and permissions

botts.ai uses a role-based access model with exactly four roles. There is one owner per organization.

A few important nuances:

• Builders cannot manage deployments or integrations. They can view deployments but cannot create or configure them — that is owner/admin only.
• Members are chat-only. A member can only use internal AI chat. They cannot view agents, knowledge bases, conversations, deployments, tools, forms, people, usage, or billing. When a member signs in they see a full-screen chat with no sidebar and no header, and any attempt to open a build or analytics page sends them back to their chat. The same boundary is enforced by the API, not just the UI.
• Internal chat oversight is owner/admin only. Owners and admins can review other members' internal AI chats; builders cannot, because member-to-AI conversations are treated as more sensitive than agent analytics. Whether the message content (not just metadata) is visible is additionally controlled per-conversation by your Internal Chat Visibility setting.

Choosing a role

Inviting members

1. Go to My Organization → Members.
2. Click Invite.
3. Enter the person's email address and choose a role: Admin, Builder, or Member. (Owner cannot be assigned by invite.)
4. Optionally select invite-only deployments to associate with the new member (these connect them to specific internal chat or phone deployments when they accept).
5. Send the invite. They receive an email with a link to join.

Invite rules:

• Seat limit. Inviting is blocked once you reach your plan's member limit. You'll be told the current count and limit and prompted to upgrade.
• Expiry. Invitations expire 72 hours after they're created.
• No duplicates. You cannot invite an email that is already a member or that already has a pending, unexpired invite.
• Pending invites appear in the Members area and can be revoked. You can't revoke an invite that has already been accepted.

Inviting and revoking invites requires the owner or admin role.

Member states, reactivation, and removal

Each member is either active or deactivated:

• Soft-deactivation is automatic. There is no manual "deactivate" button. When a plan downgrade drops you below your new seat count, members over the cap are deactivated automatically. Deactivated rows show a deactivated badge, and a deactivated user who signs in sees a screen telling them their organization access is blocked.
• Reactivate (owner/admin) flips a member back to active. Reactivation consumes a seat, so it is blocked if you're already at your member limit — upgrade first.
• Remove (owner/admin) deletes the membership. You cannot remove yourself, and you cannot remove an owner.

Changing roles and ownership

• An owner can set any role. An admin can set any role except owner.
• You cannot change your own role, and you cannot change another owner's role.
• Roles cannot be set to owner through the normal Members page.
• Ownership is single-owner. Transferring ownership is handled by botts.ai support: promoting someone to owner atomically demotes the current owner to admin, and the sole owner can never be demoted or removed. Contact us if you need to transfer ownership.

Plans and credits

botts.ai bills usage in credits. Your plan sets a monthly credit allowance plus resource limits; if you need more credits within a cycle you buy one-time packages or enable auto-recharge.

Plans

There are five tiers: Free, Lite, Standard, Pro, and Ultra. New organizations start on Free. The Plans page shows cards for Lite, Standard, Pro, and Ultra — Free isn't a card (moving down to Free is handled as a cancellation). Ultra is custom and negotiated with sales; you can't self-checkout to Ultra.

Monthly prices

The same number is used for CHF, EUR, and USD. Yearly billing is about 17% cheaper (the Plans page shows a -17% badge on the Yearly toggle).

Plan limits

Unlimited means no cap. One knowledge-base page = 2,000 characters of extracted text. Upload MB is the combined size of original uploaded files.

How credits work

Your credit balance is the sum of credits remaining across all of your non-expired credit blocks. Blocks come from different sources and expire differently:

Usage across all channels — chat, voice, widget, realtime, and internal chat — draws down this balance. Model rates are shown per 1,000 words in the model pickers.

When credits run out

If your balance reaches zero and auto-recharge is off, your agents stop responding and end-users see the message "This assistant is currently offline." This zero-credit stop is enforced server-side on every channel. With auto-recharge on, your organization is topped up before it reaches zero.

Credit packages (top-ups)

One-time credit packages let you add credits without changing plans. Packages are available on paid plans only — on Free, attempting to buy a package prompts you to upgrade first.

Each package lists its credit amount and price in your preferred currency (CHF, EUR, or USD). Buying:

• Quick purchase charges your saved card directly through a Stripe invoice. You confirm credits and amount, then pay.
• If you have no saved payment method, the flow falls back to an embedded Stripe checkout.
• Purchased credits are granted only once the payment is actually collected.

Auto-recharge

Auto-recharge automatically buys a credit package when your balance drops below 10% of your plan's monthly allowance. It is an owner/admin setting and uses a deliberate two-step flow:

1. Pick a recharge package. The auto-recharge toggle stays disabled until a package is selected.
2. Turn the toggle on. Selecting a package does not enable auto-recharge by itself — you must explicitly switch it on.

Other behavior:

• Clearing the package selection turns auto-recharge off.
• Turning auto-recharge off is always allowed.
• Re-enabling resets the failure counter.

Changing your plan

Plan changes are managed on the Plans page by owners and admins.

• Upgrades to an existing paid plan are instant and prorated. You see a preview of the change, then confirm.
• New subscriptions (from Free to a paid plan) use an embedded Stripe checkout and apply as soon as you return, so you see your new plan immediately.
• Downgrades take effect at the end of the billing period. Because a lower plan has fewer seats and agents, you choose which agents and members to keep; resources over the new cap are deactivated when the downgrade applies. Moving down to Free is handled as a cancellation.
• Cancellation also takes effect at period end, with a keep-list bounded by Free-plan limits. You can undo a pending cancellation with Keep My Plan.
• Yearly ↔ monthly: switching yearly → monthly defers to the end of your prepaid term; monthly → yearly applies immediately.

On a pending downgrade or cancellation you keep your current features and credits until the period ends, and no refund is issued. The Plans page shows a banner with the date your plan will change.

Usage

The My Organization → Usage page (owner/admin) shows how your credits are being spent:

• Credit Balance hero — your current balance, shown red when you're out of credits, amber when low (under 100), and green otherwise.
• Summary cards — high-level usage totals.
• Usage chart — a timeseries over a period you select (1–365 days), viewable as credits or tokens.
• Per-agent table — usage broken down by agent, channel, and category.
• Recent activity — individual usage events with event type, quantity, credits deducted, and time.

Builders can see the summary balance and limits (the same data behind the limit notices on build pages), but the detailed timeseries, events, and per-agent breakdowns are owner/admin only.

Data residency and privacy

You control where your AI processing happens with the Server Location setting on My Organization → General (owner/admin only). It restricts which AI models your organization can use, based on where a model is hosted and which company operates it.

There are five tiers, from strictest to most permissive:

How it's applied:

• Default is No restriction (global). Any missing or invalid value safely resolves to global, and an invalid value is rejected before it can be saved — so it can never silently loosen.
• The setting filters the model list, rejects non-compliant models when you try to select one, and flags existing deployments whose model no longer complies after you tighten the tier.
• A model whose hosting or provider country is unknown never satisfies a restrictive tier.

Privacy commitments

• Swiss platform infrastructure — platform infrastructure, chat storage, and knowledge-base storage are hosted in Switzerland (voice telephony uses a US-based provider).
• No training on your data — botts.ai does not train AI models on your content.
• Encryption — data is encrypted in transit (TLS) and at rest.
• GDPR-aware — built with European data protection regulations in mind.

For details, see our Privacy Policy and Terms of Service.

Last updated on June 16, 2026