Prompt & jailbreak testing
from
AED 16,000
Direct and indirect injection, tested and ranked.
- Injection & jailbreak test suite
- Severity-ranked findings
- Fix recommendations
- Retest included
- 2–3 week delivery
Red-team your models like you mean it.
Prompt injection, data leakage, and jailbreak testing before an attacker finds them first.
What's included
Everything in the base engagement.
No upsell tiers. Every deliverable ships in the base price you sign for.
8
Get the full scope in writingdeliverables, zero add-ons
all in the quote you sign
Base engagement
Discovery workshop
systems mapped to actual risk
Working assessment
against your real deployments
Production-grade report
evidence, not just opinions
Executive readout
plain-English risk summary
Remediation roadmap
prioritized by risk and effort
Data privacy review
your data stays yours
Documentation & handover
policies, controls, and evidence
2 weeks of follow-up
included after delivery
Types we build
Pick the audit you need.
Every level digs deeper into how your models can be broken. Select one to see what each involves.
Prompt injection & jailbreak testing
Typical timeline
2–3 weeks
Typical coverage
One model/app
What's specific to this type
- Direct & indirect injection tests
- Jailbreak attempt library
- Severity-ranked findings
How it goes
Four steps, no black box.
Scroll to step through the engagement ↓
Step 01 / 04
Kickoff call
A focused call to lock the scope, systems in play, and a fixed quote. No open-ended discovery retainer. You leave with a written scope before anything starts.
You get
a written scope and a fixed quote
It starts with one call.
Thirty minutes to a scope, a quote, and a fixed start date.
Built on
What we reach for red-teaming your models.
PythonHugging FaceLangChainOpenAIAnthropicDatadog
Pricing
Starting prices, in the open.
Real numbers, not "contact us for pricing." Your final quote depends on scope and number of systems. You'll have it in writing within 48 hours.
Data leakage assessment
from
AED 32,000
Where your model might be leaking data it shouldn't.
- Training data extraction tests
- PII & secrets leakage checks
- Output filtering review
- Remediation plan
- 3–4 week delivery
Full adversarial audit
from
AED 70,000
Every attack surface, tested, with a regression gate.
- Injection, leakage & jailbreaks
- Custom red-team harness
- CI/CD regression gate
- Priority support
- 5–7 week delivery
What we can build
What the audit actually tests.
Illustrative concepts, not real client projects, just a quick read on how we approach this.
Jailbreak test suite
A living library of attempts, run against every release.
Data exfiltration testing
Checks whether your model gives up what it shouldn't.
Regression security gate
Blocks a release if a known exploit resurfaces.
Questions
Model Security Auditing, answered.
Yes. We assess third-party and vendor AI systems the same way, you don't need to have built it in-house for us to review it.
Most engagements are a one-time assessment or fixed retainer, no per-API-call costs like a build project. We size the retainer with you before we start, no surprise bill.
Prompt injection testing in 2–3 weeks, a full adversarial audit in 5–7.
Yes. Findings go to you first, with a fix, before anything is written up for a broader audience, if you choose to publish at all.
Yes, the CI/CD gate option re-runs the core test suite on every deploy and blocks the release if a known exploit resurfaces.
A clear, proven process
Model audits.
Evidence to ship.
Prompt, data, weights, deploy. Fixed quote in 48 hours, audit pack the same sprint.
The processbrief → first staging build in ~2 weeks
01Send the briefWhat you’re building, in a paragraph.5 min
02Get a fixed quoteScope, price, and date, locked.48 hrs
03Watch it take shapeA new build to staging every week.weekly
04Ship to productionLive, handed over, and supported.live
team onlineavg reply 1h 47m