Pricing Options
Find the right tier for your needs.
Basic
For teams in need of a simple
prompt hacking test suite.
First month: $450
On-going: $200 / month
Select this Package
✓ Prompt hacking tests starts within 5 business days
✓ Level 2 prompt hacking test suite (Simple, Direct, and some Advanced Prompt Injections)
✓ 1 product area
✓ Best practice methodology
✓ Standard testing parameters
✓ Detailed findings and recommended fixes
✓ Async reporting via email
Essentials
For teams looking to gain
confidence in their AI security.
First month: $800
On-going: $350 / month
Select this Package
✓ Prompt hacking tests start within 3 business days
✓ Level 5 prompt hacking test suite (Basic, Simple, Advanced, Indirect, Jailbreaking, etc)
✓ 1 product area
✓ Best practice methodology
✓ Standard testing parameters
✓ Detailed findings and recommended fixes
✓ Real-time collaboration via Slack or Discord
Premium
For teams that want on-going robust AI security coverage.
Contact Sales
Get a Quote
✓ Prompt hacking tests start within 1 business day
✓ Unlimited runs of our level 5 prompt hacking test suite (Basic, Simple, Advanced, Indirect, Jailbreaking, etc)
✓ Unlimited product areas
✓ Best practice methodology
✓ Ability for custom testing parameters (Geo, Time Zone, Testing Windows)
✓ Detailed findings and recommended fixes
✓ Real-time collaboration with your chat product of choice + Zoom calls when critical
✓ Consultation on AI safety, governance, and controls
✓ AI strategy sessions
Frequently Asked Questions
What is the process like?
Primarily async and swift.
Once payment has been confirmed, we'll send a form for you to fill out. This will help us understand your product, give us the right access, and prepare the right tests. Once done, we can begin testing promptly.
Why primarily async?
We know you're busy.
No one benefits from being stacked with back-to-back hour long sync meetings. We deliver results primarily async and only do video calls when absolutely necessary.
How does the 'First month' and 'On-going' pricing work?
In the first month, we run the latest version of our test suite and report findings with recommended fixes. If the fix has been patched within 30 days, we will re-test at no additional cost.
Stay secure from the latest attack methods with our monthly plan. We'll run the latest version of our test suite and report findings.
What do the different levels of prompt hacking mean?
It covers the different levels of prompt hacking techniques.
The test suite is ever evolving as the industry changes: new models are released, new prompt hacking techniques are discovered, and new threats emerge. As the landscape shifts, so will our test suites we offer.
Where do your methodologies come from?
Our own products and hundreds of hours researching in ARXIV.
After building a few Generative AI products since 2023, we learned how difficult it is to handle the non-deterministic nature of LLMs. Now we're bringing our hours of creating our own products and researching adversarial prompt hacking methods to bring them to others.
How can I be 100% certain to never have an AI security risk?
You can't. You can only mitigate currently known attacks.
Generative AI security is in 'beta' at the moment. In February of 2024, the United States announced the First-Ever Consortium Dedicated to AI Safety. This is a developing field and we're here to equip you with the latest knowledge to protect your data, your business, and your reputation.
Get in touch
Still have a question? Leave a message.
Protect your business, secure your data, and enable growth with Model Security.
© 2024 Model Security. All rights reserved.
We need your consent to load the translations
We use a third-party service to translate the website content that may collect data about your activity. Please review the details and accept the service to view the translations.