Pick a plan by team size. Upgrade anytime with prorated billing, or downgrade at the end of a cycle.
For one person running a pilot.
For teams past their first pilot.
For multiple teams that need SSO and compliance.
Extra testers are billed per tester, with an optional cap.
| Feature | Starter | Growth | Scale |
|---|---|---|---|
| Admin seats | 2 | 5 | 15 |
| Annotators | 2 | 8 | 25 |
| Tester pool | 15 | 50 | 200 |
| Active projects | 3 | 10 | Unlimited |
| Rubric + pairwise | ✓ | ✓ | ✓ |
| Context layer + API | ✓ | ✓ | ✓ |
| Chat / quiz / media / API / collection | - | ✓ | ✓ |
| Feedback API + webhook import | - | ✓ | ✓ |
| White-label + SSO | - | - | ✓ |
| Custom data residency | - | - | ✓ |
Yes. 14 days of the full product on any tier, no card required.
Each invited person with the tester role. Annotators and admins are counted separately.
Add them anytime. Extra testers are prorated for the days left in the cycle.
No. Evaluations are unlimited on every tier. You pay only for seats and testers.
Yes. Your own API, Ollama, or any OpenAI-compatible endpoint.
Rules and examples drawn from your reviewed feedback. You approve each one, and your model reads them through one API call.
You can export until the billing period ends. We keep your data 30 more days, then delete it permanently.