Phase
Develop
Run local CLI scans while building agent logic and tool integrations.
Simple Pricing
Security At Every Stage
Start free while building, then scale as agent workloads grow. All plans include full threat taxonomy coverage.
Phase
Integrate
Add adversarial testing in CI/CD before releasing agent updates.
Phase
Operate
Monitor production behavior and enforce runtime security posture.
FREE
Scan
₹0
Also available as $0/mo
Full adversarial scan, local execution, no account required.
1,000 attack vectors / month (~4 full scans)
CLI security testing
Single repository
Prompt injection detection
Agentic Top 10 threats coverage
TEAM
Go
₹2,499/mo
Also available as $29/mo
For teams shipping AI agents into production environments.
20,000 attack vectors / month (~90 scans)
Up to 3 repositories
CI/CD security testing
Extended adversarial attack packs
Agent vulnerability coverage
Scan history & security reports
MOST POPULAR
ENTERPRISE
Scale
₹7,999/mo
Also available as $99/mo
Advanced runtime monitoring, telemetry, and enterprise controls.
100,000 attack vectors / month (~450 scans)
Unlimited repositories
Production runtime monitoring
Agent execution telemetry
Security posture scoring
Continuous vulnerability detection
Comparison
Feature Comparison
Choose the plan that fits your AI agent security requirements.
Scan
1,000 vectors/mo
- ✓1,000 attack vectors / month (~4 full scans)
- ✓CLI security testing
- ✓Single repository
- ✓Prompt injection detection
- ✓Agentic Top 10 threats coverage
- ✓Structured vulnerability reports
Go
20,000 vectors/mo
- ✓20,000 attack vectors / month (~90 scans)
- ✓Up to 3 repositories
- ✓CI/CD security testing
- ✓Extended adversarial attack packs
- ✓Agent vulnerability coverage
- ✓Scan history & security reports
Scale
100,000 vectors/mo
- ✓100,000 attack vectors / month (~450 scans)
- ✓Unlimited repositories
- ✓Production runtime monitoring
- ✓Agent execution telemetry
- ✓Security posture scoring
- ✓Continuous vulnerability detection
- ✓Priority support
Common Questions
What counts as a scan?
Each call to the FortifAI scan endpoint counts as one scan event. A typical `npx fortifai scan` run usually maps to one billable scan call.
Can I switch plans later?
Yes. Upgrades apply immediately. Downgrades apply at the end of the current billing cycle.
Is data sent to third-party models?
No. FortifAI uses deterministic security evaluation workflows and does not forward your payloads to secondary LLMs for classification.
Which payment methods are supported?
Stripe supports card and wallet flows. Razorpay supports UPI, cards, net banking, and wallets based on region availability.
What happens when quota is exceeded?
Requests are blocked with a clear quota response so your team can upgrade or wait for quota reset. There are no silent drops.
Start Securing Your AI Agents
Detect prompt injection, tool abuse, and data leakage before risks reach production systems.
No credit card required.