Open models keep inference cheap, so you get more for less. Start free and upgrade when your customers do the talking.
| Feature | Free | Pro | Business | Enterprise |
|---|---|---|---|---|
| Team seats | 1 | 3 | 10 | Unlimited |
| Bots | 1 | 3 | 10 | Unlimited |
| Messages / month | 100 | 2,000 | 10,000 | Unlimited |
| Knowledge (characters) | 500K | 10M | 100M | Unlimited |
| Documents & website ingest | ||||
| Hybrid retrieval & citations | ||||
| Live SQL & API data | ||||
| Live data connections | 2 | 10 | Unlimited | |
| Live data requests / month | 20,000 | 100,000 | Unlimited | |
| Live visitor tracking | ||||
| Visitor analytics history | 7 days | 90 days | 365 days | 730 days |
| Campaigns, journeys & conversion funnel | ||||
| Slack & WhatsApp channels | ||||
| Remove branding | ||||
| White-label |
No. The Free plan is genuinely free — one bot and 100 messages a month, no card required. Add a card only when you upgrade.
Each answered question from a visitor counts as one message. Cached answers still count, but cost you nothing in inference.
One executed query or API call against your own systems — whether your bot made it to answer a question, or someone loaded a dashboard built on it. Repeated dashboard views served from cache don't count, because nothing ran against your database.
Never. We run open models via OpenRouter under our control; your content is used only to answer your users' questions.
Connections are read-only and encrypted (AES-256). Queries run in a read-only transaction with a statement timeout, and every call is audit-logged.
Yes — upgrade or downgrade anytime from your workspace settings. Changes take effect immediately.
Create a free workspace and have your first bot answering in minutes.