F
just run that shit
LLM FREEDOM AUDITS
TRANSPARENT PRICING • NO BULLSHIT

Pay for real private audits.
Not theater.

Free tier for demos and public replication. Paid tiers for production evidence using your models and keys. Results stay yours.

FREE
Public Demo
$0
Instant • No key required
  • • Heuristic simulator (no real LLM calls)
  • • 6 scenarios max per run
  • • Public results (demo only)
  • • Basic JSON export
  • • See the 40% DeepSeek report
RUN FREE DEMO
Great for understanding the instrument
ENTERPRISE
Unlimited + White Label
$999/mo
Per month • Teams & Labs (annual billing on request)
  • • Unlimited private audits (any volume)
  • • Your models + our keys or bring your own
  • • Custom templates & scenario generators
  • • White-label reports (your logo, NDA options)
  • • Private team dashboard + API + history
  • • Quarterly 30-min consulting review
  • • Priority queue + dedicated support
  • • On-prem / airgap deployment available
START 14-DAY ENTERPRISE TRIAL
Includes Manifund-style open research credits for qualifying labs
SECURE CHECKOUT • POWERED BY STRIPE

Ready for a private audit?

Both tiers live on Stripe. Click below to go to secure hosted Checkout — we email you the private run link + key upload form (or Enterprise onboarding) right after payment. Results stay private to you.

PAY WITH STRIPE — $199 ONE-TIME (Private Production Run) SUBSCRIBE — $999/mo ENTERPRISE (Unlimited + White Label)
Cards, Apple Pay, Google Pay • Secure hosted checkout
Prefer invoice / wire / crypto? Email us
Enterprise POs accepted. Net-30 available for approved labs.
What do I get for $199?
A complete paired audit (observed vs max-private framing) on up to 30 dilemmas using the exact methodology that produced the public 40% DeepSeek result. Your exact prompts + model responses (raw), action matrix, freedom-delta calculations, and a polished report. Private by default — we do not publish or train on it.
How private is "private"?
For one-off: we process with your key ephemerally, delete after report generation unless you opt into 30-day dashboard. Enterprise: full data residency control + contract. We never log your prompts for training or resale.
Can I use my own models / endpoints?
Yes. Any litellm-supported provider (OpenAI, Anthropic, DeepSeek, Grok, local via vLLM, Together, Fireworks, etc.). Bring your key. We support custom base URLs for private clusters.
Enterprise vs one-off?
One-off is perfect for a single high-stakes model release audit or paper. Enterprise is for continuous red-teaming, eval harness integration, multiple teams, custom scenario libraries, and white-labeled deliverables for regulators or investors.
Questions? tim@justrunthatshit.com • See full Terms + Liability Disclaimer