Home / Pricing
Simple, Transparent Pricing
Start free. Scale as your evaluation needs grow. No surprise fees, no hidden per-model charges.
Solo & Exploration
Free forever for individual developers evaluating a single model. No credit card required.
- 1 model endpoint
- 500 evaluations / month
- 5 test suite presets
- Hallucination detection (basic)
- Python and Node.js SDK
- Community support
- 7-day evaluation history
- Red-teaming engine
- CI/CD integration
- Custom metrics
Growing AI Teams
Everything a product AI team needs to ship with confidence. Billed annually.
- Up to 10 model endpoints
- 25,000 evaluations / month
- Full test suite library (200+)
- Hallucination detection (advanced)
- Red-teaming engine (50 scenarios)
- GitHub Actions + GitLab CI
- Custom scoring rubrics
- 90-day evaluation history
- Email + Slack support
- Up to 5 team seats
Scale and Compliance
Unlimited evaluations, dedicated infrastructure, and enterprise-grade compliance for regulated industries and large deployments.
- Unlimited model endpoints
- Unlimited evaluations
- Full red-teaming library (150+)
- Custom hallucination taxonomy
- All CI/CD integrations
- SSO / SAML
- SOC 2 Type II reports
- Dedicated account manager
- SLA 99.9% uptime
- On-premise deployment option
All plans include standard security features. Team and Enterprise plans qualify for annual billing discounts. Contact us for non-profit or academic pricing.
Plan Comparison
| Feature | Developer | Team | Enterprise |
|---|---|---|---|
| Evaluations / month | 500 | 25,000 | Unlimited |
| Model endpoints | 1 | 10 | Unlimited |
| Hallucination detection | Basic | Advanced | Custom |
| Red-teaming engine | 50 scenarios | 150+ scenarios | |
| CI/CD integration | |||
| SSO / SAML | |||
| Support | Community | Email + Slack | Dedicated CSM |
Pricing FAQ
What counts as an evaluation?
One evaluation is a single LLM output assessed against your test criteria. Running a 100-case test suite on one model version counts as 100 evaluations. Batch runs are counted per individual output assessed.
Can I switch plans mid-cycle?
Yes. Upgrades take effect immediately and are prorated. Downgrades take effect at the start of your next billing cycle. No penalties for changing plans.
Do you offer annual billing discounts?
Yes. Annual billing on the Team plan saves 20% compared to monthly. Enterprise contracts include custom annual pricing. Contact us for details.
What models does Confident AI support?
All plans support OpenAI, Anthropic, Mistral, Meta Llama, Google Gemini, and any custom model accessible via REST API or our SDK. Model count limits apply to simultaneously configured endpoints, not to which providers you use.
Is there a free trial for the Team plan?
Yes. The Team plan comes with a 30-day free trial, no credit card required. Your trial includes full access to all Team features including the red-teaming engine and CI/CD integrations.
Not Sure Which Plan Is Right?
Our team can walk you through which plan fits your evaluation volume, team size, and compliance requirements. Most conversations take under 20 minutes.