Start free. Scale when ready.

500 evaluations a month — shared across up to 2 agents — is enough to experience the system fully. The verified path and the reliability-tier credential unlock when you're ready for production.

Free
$0forever

For developers and researchers exploring HB-Eval.

Start free
Up to 2 agents
500 evaluations / month (shared across your agents)
All 5 reliability metrics (PEI, FRR, IRS, TI, CSI)
Fault-injection battery (6 fault types × 6 domains)
Local evaluation path (unverified results)
Full diagnostic guidance (why you failed + how to fix)
EDM memory — 3 retrievals
Full SDK access (hb-eval-sdk)
Live dashboard
30-day evaluation history
Community support (GitHub)
Verified evaluation path (platform-run, tamper-proof)
Reliability-tier credential (Meets Tier 1/2/3)
HCI-EDM performance-grounded explanations
Unlimited memory retrieval
Most Popular
Pro
$49per month

For teams running agents in production who need verified, tier-qualified reliability.

Join waitlist
Up to 10 agents
5,000 evaluations / month (shared across your agents)
All 5 reliability metrics (PEI, FRR, IRS, TI, CSI)
Fault-injection battery (6 fault types × 6 domains)
Verified evaluation path — platform-run, tamper-proof
Reliability-tier credential — Meets Tier 1 / 2 / 3
HCI-EDM explanations (grounded in your stored episodes)
Unlimited EDM memory retrieval
Full diagnostic guidance
Full SDK access
Live dashboard
12-month evaluation history
Priority support (email)
Enterprise
Customcontact us

For organizations that need custom limits, SLAs, and onboarding.

Contact us
Custom agent limit
Custom monthly evaluation volume
All Pro features
Verified path + all reliability tiers (1 / 2 / 3)
Custom integration support
Dedicated onboarding
Audit-log export
SLA guarantee

Pro plan is coming soon

We're rolling out Pro to a small group first. Leave your email and we'll reach out before the public launch — early users get a founding-member discount.

Join the waitlist