Example system pricing — AI excluded
You pay for the platform — not for every second an agent idles in the background. All prices apply only to the ThinkShip system license: hosting support, updates and support. These are illustrative example prices, not a binding offer.
System and AI costs are separate
These prices cover only hosting support, updates and support for the ThinkShip platform. AI model costs (cloud API or local hardware) are not included — each tenant uses its own provider key or its own local models and pays for that directly, or bears its own power costs. There is no shared cloud account — your data, your sovereignty. And because recurring checks (deadlines, stock, delays) run as a workflow instead of an agent running around the clock, the monitoring itself doesn't rack up an ongoing AI bill in the first place — only the moment something actually needs a decision does.
For scale: always-on vs. workflow-first
An AI agent checking a single watch (e.g. a deadline check) hourly via a language model: 24 checks × 30 days × roughly 800–1,200 tokens per check.
The same check runs as a cron job for 0 tokens. AI only switches in on the days a real hit occurs — typically 2–8 hits a month for a monitoring task × roughly 1,000–2,000 tokens.
Illustrative figures for orientation, not a guarantee — actual token counts depend on the model, check frequency, context length and prompt. For a single example, not your whole operation.
Team
For small teams who want to self-host ThinkShip.
- Up to 10 users
- Up to 5 concurrently active agent roles
- Self-hosting guide & installer
- Community support
Business
For organizations running ThinkShip across multiple departments.
- Up to 50 users
- Unlimited agent roles
- All departments & compliance features
- Priority email support
- Update service
Enterprise
For large organizations with individual requirements.
- Unlimited users
- Dedicated onboarding & SLA
- On-premise / air-gapped option
- Custom integrations
Side by side
| Feature | Team | Business | Enterprise |
|---|---|---|---|
| Users | up to 10 | up to 50 | unlimited |
| Concurrently active agent roles | up to 5 | unlimited | unlimited |
| Departments / use cases | selected | all | all |
| Self-hosting support | guide | guide + support | guided onboarding |
| On-premise / air-gapped | – | – | ✓ |
| Support channel | community | priority (email) | dedicated + SLA |
Pricing questions
What does the AI itself cost?
Not part of these system prices. Each tenant stores its own encrypted provider key (e.g. for a cloud model) or runs its own local models. Actual AI costs depend on the provider, model size and usage volume, and are billed directly between you and your chosen provider — for purely local operation, they show up as your own hardware cost instead.
Are these real market prices?
No, these are illustrative example prices for orientation only, net of tax and without guarantee. An individual quote depends on actual user and usage numbers.