Pricing
Start free.
Scale when you do.
The proof layer is free and stays free. 1,000 anchored sessions a month, no credit card. Pay only when you want the evidence escrowed, retained, and provably destroyed.
Evidence
Free
1,000 anchored sessions / month
- Local evidence, public anchors
- Verify URLs anyone can check
- The proof layer is free and stays free
Retain
$29
/ month
- Everything in Evidence
- Encrypted escrow of every session
- 1-year retention
- Recover from any machine with one file
- For anyone whose laptop should not be the archive
Recommended
Comply
$349
/ month
- 7-year retention
- Attested destruction at expiry
- A deletion certificate per session
- Retention report endpoint for auditors
- For teams whose auditors ask where agent records live
Agents
$0.05
/ session, x402
- 7-year retention
- No account
- Paid in USDC by the agent itself
- Live today on Base
- Your agent can buy its own evidence
Enterprise
By email
Custom terms
- 10-year retention
- WORM language
- Custom terms
- Write to scott@rubric-protocol.com
Every tier: your key, our ciphertext. Rubric cannot read your evidence, and recovery never requires our cooperation. That is not a policy. It is the architecture. PRICING.md →
Design Partner Program
Get examination-ready before
the standards land.
We take on a small cohort of regulated AI teams who need to meet EU AI Act obligations (high-risk from December 2027) and US state law obligations already in force. You get full Enterprise access free for 60 days. We get a reference customer and a joint case study.
01
60-day free pilot
Full Enterprise tier. No credit card. No commitment.
02
$2,500 / month after pilot
Locked in — Enterprise list price is $9,999 / month.
03
Co-authored case study
Your logo, your story. Published when you are ready.
04
Direct line to the founder
Integration support, custom compliance mapping, priority roadmap input.
Compliance Coverage
Global AI compliance infrastructure. One platform, every major framework.
One platform. Every framework your examiners will ask about.
| Framework | Rubric Control | Available On |
|---|---|---|
| EU AI Act Art. 12 | Automatic logging of AI decisions — tamper-proof, timestamped, independently verifiable | All tiers |
| EU AI Act Art. 14 | Human oversight and override evidence — attested chain of custody from AI output to human sign-off | Enterprise+ |
| GDPR Art. 22 | Automated decision documentation — cryptographic proof retrievable for data subject requests | Enterprise+ |
| SR 26-2 | AI model inventory with SR 26-2 risk tiers, validation status, outcomes monitoring — auto-built from attestation data | Enterprise+ |
| FINRA 3110/3120 | Supervision records and attestation reports — examination-ready evidence package in a single API call | Enterprise+ |
| FFIEC IT Exam | Third-party AI vendor audit completeness, access log evidence, policy enforcement summary | Enterprise+ |
| OCC 2023 AI | Delegated authority and agent identity trails — cryptographic binding of AI agent to human principal | Dedicated |
How Rubric Fits
Not a replacement.
The missing layer.
vs. Internal Logging (Splunk, Datadog)
Internal logs are self-certified — they live in your stack, controlled by your team. Rubric anchors attestations to an independent, immutable ledger. Internal integrity hashes are generated and held by the same party that controls the logs. When your examiner asks for proof the log was not altered, Rubric's answer comes from an independent ledger.
vs. AI Governance Platforms
AI governance platforms generate examination reports from their own data — self-certified by design. Rubric is the independent cryptographic anchor underneath. Use both: governance tooling for policy enforcement, Rubric for immutable proof.
vs. Manual Audit Prep
A regulatory examination cycle costs $500K–$2M in consulting time. Rubric generates your examination-ready package in a single API call. The audit infrastructure pays for itself on the first examination.