Performance Data
TurboPentest Benchmarks
Transparent performance data. What TurboPentest runs, what it covers, and how it compares - with formal benchmark results coming as we complete testing.
Last updated: June 2026
14
Security Tools
Orchestrated by Paladin AI
10/10
OWASP Top 10 Coverage
2025 edition
A few
Hours to Report
Audit-Ready tier
$99
Per Pentest
Audit-Ready tier
OWASP Top 10 2025 Coverage
TurboPentest targets all 10 categories of the OWASP Top 10 2025 through a combination of specialized security tools and Paladin AI analysis. See our full testing methodology for how PTES, OWASP, and MITRE ATT&CK map across the pentest.
OWASP LLM Top 10 - Coverage Scope
Paladin runs with current OWASP testing guides - including the LLM Top 10 - loaded as agent context. Every category is in scope. Coverage quality is bounded by what the frontier models powering Paladin can attempt at the time of the test. As frontier-model capabilities improve, so does our coverage depth.
We don't publish a per-category pass/fail table because the underlying capability is dynamic, not a fixed feature matrix. What we commit to: we attempt every category within your configured pentest scope, we reason about application behavior the same way a human pentester would, and we surface findings honestly - including limitations in what we could attempt during the run.
The 10 categories Paladin attempts
Need a worked example? See /for/ai-companies for a synthesized illustrative AI finding.
Speed and Cost Comparison
How TurboPentest compares to traditional approaches in time and cost.
14-Tool Security Stack
Every pentest orchestrates 14 industry-standard security tools in parallel, validated by Paladin AI. Black-box tools run on every pentest. White-box tools activate when you connect your GitHub repository.
Port Scanner
Network
Web Scanner
Web App
Vuln Scanner
Multi-vector
Server Audit
Web Server
TLS Analyzer
SSL/TLS
Sub Hunter
Subdomain
Enumerator
Fuzzing
Web Probe
Fingerprinting
WAF Detect
WAF Detection
Net Scanner
Vulnerability
Security Checks
Custom Checks
Code Scanner
SAST
Dep Scanner
SCA
Secret Scanner
Secrets
Paladin AI
Validation
Formal Benchmarks In Progress
We're running TurboPentest against recognized industry benchmarks and will publish results here as each test completes. No scores until we've earned them.
XBOW Validation Benchmark
104 CTF-style challenges mirroring real-world vulnerability classes. The emerging standard for evaluating agentic pentesting agents.
View benchmark source →Vulhub CVE Speed Benchmark
Timed detection and exploitation of 30 real-world CVEs from Vulhub, compared against manual pentesting and standalone scanners.
View benchmark source →OWASP Benchmark Scorecard
Formal scorecard against the OWASP Benchmark Project, measuring true positive and false positive rates across thousands of test cases.
View benchmark source →Our Approach to Benchmarking
We believe in transparent, reproducible benchmarks. No scores we haven't tested. No claims we can't back up.
Public Benchmarks Only
We test against publicly available benchmarks so results are independently verifiable. No proprietary test suites designed to make us look good.
Results After Testing
We publish scores only after completing each benchmark ourselves. No projections, no estimates, no borrowed numbers from component tools.
Continuous Re-testing
We re-run benchmarks with every major Paladin update and publish updated scores. This page always reflects the latest verified results.
See the Results Yourself
Run a professional agentic pentest on your domain. 14 tools, Paladin AI validation, results in a few hours.