Free for 50,000 traces a month. Then $49.
Every plan ships the full eval engine, all 300+ attack plugins, and the AI Gateway. Upgrades add traces, seats, retention, and the team + compliance toolset.
Billed monthly in USD. Cancel any time. Annual invoicing on request.
Free
More generous than Langfuse, Helicone, Portkey, LangSmith — combined.
- 1 team members
- 50,000 traces/month
- 100 red team scans/month
- Unlimited projects
- 30-day data retention
- All 200+ built-in scorers + custom
- All 300+ attack plugins
- Full 5-layer firewall protection
- AI Gateway (route, cache, failover) — 10,000 requests/month
- Pairwise eval & ELO leaderboard
- Annotation queues
- CI/CD integration
- OTel-native observability
- Discord + community support
Pro
Everything in Free, plus production scale, FinOps and the Prompt IDE.
- Unlimited team members
- 500,000 traces/month
- Unlimited red team scans
- Unlimited projects
- 90-day data retention
- All 200+ scorers + custom scorers
- AI Gateway (route, cache, failover) — 1,000,000 requests/month
- FinOps dashboard
- Prompt IDE (version, A/B test, optimizer)
- Full 5-layer firewall protection
- Annotation queues
- Pairwise eval & ELO leaderboard
- Email support
Team
Everything in Pro, plus the governance and compliance toolset.
- Unlimited team members
- 2,000,000 traces/month
- 365-day data retention
- RBAC & SSO/SAML
- 50 compliance frameworks
- SOC 2 readiness reports
- Audit logging
- Custom webhooks & dashboards
- Data residency controls
- MCP server registry & governance
- Priority support
Enterprise
Custom deployment with unlimited scale and dedicated support.
- Unlimited members
- Unlimited traces
- Custom data retention
- VPC/on-premise deployment
- Custom SLAs (99.95%)
- GDPR DPA
- Dedicated account manager
- 24/7 phone support
- Custom integrations
- Training & onboarding
Estimate your cost
Adjust the sliders to find the right plan for your usage.
Recommended plan
Free at $0
In the free tier
The whole platform — before you pay a cent.
Most tools gate their best capabilities behind a paid plan. EvalGuard ships the full eval engine, every attack plugin, and the complete 5-layer firewall on the free tier. Every number below is drift-checked against our registry in CI.
How 200+ scorers and 300+ attack plugins stack up against DeepEval, Promptfoo, Garak, and PyRIT: see the coverage benchmarks.
Side by side
Compare all features
| Feature | Free | Pro | Team | Enterprise |
|---|---|---|---|---|
| Team members | 1 | Unlimited | Unlimited | Unlimited |
| Traces/month | 50,000 | 500,000 | 2,000,000 | Unlimited |
| Logs & metric points/month | 50,000 each | 500,000 each | 2,000,000 each | Unlimited |
| Red team scans | 100/mo | Unlimited | Unlimited | Unlimited |
| Data retention | 30 days | 90 days | 365 days | Custom |
| Projects | Unlimited | Unlimited | Unlimited | Unlimited |
| Built-in scorers | 200+ | 200+ | 200+ | 200+ |
| Custom scorers | ||||
| Attack plugins | 300+ | 300+ | 300+ | 300+ |
| AI Gateway (route/cache/failover) | 10,000 req/mo | 1,000,000 req/mo | 5,000,000 req/mo | Unlimited |
| Pairwise eval & ELO leaderboard | ||||
| Annotation queues | ||||
| FinOps dashboard | — | |||
| Prompt IDE (incl. optimizer) | — | |||
| CI/CD integration | ||||
| SSO / SAML | — | — | ||
| RBAC | — | — | ||
| Audit logging | — | — | ||
| SOC 2 readiness | — | — | ||
| Compliance frameworks | — | — | 50 | 50+ |
| Custom webhooks | — | — | ||
| VPC / on-premise | — | — | — | |
| Custom SLA | — | — | — | |
| GDPR DPA | — | — | — | |
| Dedicated account manager | — | — | — | |
| 24/7 phone support | — | — | — |
How we count. A trace is one end-to-end request — an agent run, a chain, a single LLM call. The spans inside it are not billed separately, however many your instrumentation emits. Logs and metric points have their own monthly allowances and never consume your trace allowance. Quotas reset at the start of each UTC month.
Questions
Frequently asked questions
Ready to ship AI with confidence?
Free forever for individuals. Pro plans start at $49/mo. No credit card required to get started.