Per its own site in November 2026, an autonomous offensive security platform that tests web applications and APIs and reports findings it has validated with a working exploit.

Is PentestLLM an autonomous pentester?

No. It is a model API. It answers requests from your agent or script, and the workflow, the tools and the approvals are yours.

Can I build my own pentest agent on PentestLLM?

Yes, that is the intended use, for authorized testing. Any agent that takes an OpenAI-compatible or Anthropic-compatible endpoint can use it, and the agent guides cover OpenCode, Claude Code, Cline, aider and others.

Does XBOW let me choose the model?

Its platform page says XBOW benchmarks models and routes each task itself. The pages we read in November 2026 did not describe bringing your own model.

What does PentestLLM cost?

Per token from a prepaid balance with a $10 minimum top-up by card and no contract. On 2026-11-12 input prices for security builds start at $0.40 per million tokens.

An XBOW alternative you build yourself

XBOW is an autonomous offensive security platform that runs the whole assessment for an organization; PentestLLM is not that, it is the model API for testers who run their own agent and need a security-tuned model behind it. One prepaid key from $10 reaches 9 security-tuned builds, billed per token.

Get a key Prices per token

shell

curl https://api.pentestllm.ai/v1/chat/completions \
  -H "Authorization: Bearer sk-ptl-live-…" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "auto/security",
    "messages": [{"role": "user", "content": "Rank these scanner findings by severity and say what to verify first."}]
  }'

What XBOW is, as of November 2026

XBOW’s site, read on 11 November 2026, describes an autonomous offensive security platform for AI-driven pentesting of web applications and APIs. Its platform page lays out six stages, from learning and mapping the target to attacking, proving and verifying, and says a coordinator directs agents while independent validators check that each exploit reproduces. Findings come with the request, the response and reproduction steps, and the platform retests after a fix.

The same pages describe controls for the buyer: scope, impact levels, rate limits, testing windows, a safety check on each action and a full log. On models, the platform page says there is no permanent best model and that XBOW benchmarks models and routes each task itself. It names no model and does not describe bringing your own.

Its pricing page listed no prices on the day we read it. It describes usage-based pricing scoped to your environment, sold through a demo, a quote request or the AWS, Google, Oracle and Microsoft cloud marketplaces. We have not used XBOW and repeat none of its performance claims here.

Different buyers

What you get: XBOW: a platform that runs the assessment and delivers validated findings. PentestLLM: a model endpoint for an agent you run.
Who it is for: XBOW: organizations buying continuous testing of their own applications. PentestLLM: individual testers, red teams, bug bounty hunters and CTF players.
How you buy: XBOW: a demo, a quote or a cloud marketplace, per its pricing page. PentestLLM: self-serve, prepaid from $10 by card.
Who drives: XBOW: its own agents, inside the guardrails it describes. PentestLLM: you, through your agent and your approval rules.
The model: XBOW: chosen by the platform. PentestLLM: chosen by you per request, or by the auto/security router.

These are not substitutes in the usual sense. One is a finished service and the other is a part.

Why testers want the part

A consultant, a bug bounty hunter or an internal red team usually has a method already, and often an agent: OpenCode or Claude Code in a terminal, a Python harness, or an open-source pentest agent such as Strix or PentestGPT. What that setup can lack is a model that will engage. A general model declines many offensive security questions, and one declined step can stall an agent run.

PentestLLM supplies that piece. The builds are tuned or abliterated to answer security questions other models decline, which is reduced refusals, not none.

What the model API includes

An OpenAI-compatible endpoint at https://api.pentestllm.ai/v1 and an Anthropic-compatible Messages endpoint on the same key, with context windows up to 1M tokens. The id auto/security routes each request to a fast, balanced or frontier security build and bills at that build’s price. When a request declares them, web search ($0.01 per search), page fetch ($0.001 per page fetched), a code sandbox ($0.05 per container-hour, 1-minute minimum) and remote MCP (No tool fee) run on the API side. For running tools on your own Kali machine, see the Kali MCP guide.

What it does not do

PentestLLM ships no agent and no scanner. It will not map an application, run an assessment, validate a finding or retest a fix. It has no scope controls, because it never touches a target: your agent does, under your rules. If you need those things as a product, a model API is the wrong purchase.

The cost follows from that. There is no contract, only a prepaid balance spent per token. On 2026-11-12 input prices for security builds start at $0.40 per million tokens, and the table below is the live catalog.

When XBOW is the better choice

If you are responsible for many applications and want them tested continuously without building anything, XBOW’s site describes exactly that: autonomous assessment, validated findings with proof, retesting and reporting for auditors. A security team without the time to run its own agents, or a company that needs findings it can hand to developers with reproduction steps, is its buyer. PentestLLM is for the person who wants to do the testing and needs the model.

Both come with the same condition. XBOW tests within the scope its customer sets, and PentestLLM is for systems you own or have written permission to test, per the acceptable use policy.

9 security builds, priced per token

Read from the live catalog on 2026-11-12. Prices are US dollars per million tokens, paid from a prepaid balance that starts at $10. The id auto/security has no price of its own: each request is billed at the build that answered.

Build	Input	Output	Context
Qwen3.8 27B Cybersecurity	$0.40	$2.40	262,144
Gemma 4 26B A4B Cybersecurity	$0.4224	$1.3376	262,144
GLM-5.3 Flash Cybersecurity	$0.5093	$1.6975	1,048,576
Red Team (Flash)	$0.90	$3.00	not published
Red Team (Pro)	$1.50	$6.00	not published
OrcaCyber MiniMax M3.1 Flash	$2.00	$4.00	261,632
OrcaCyber Qwen3.8 Flash Next	$2.00	$4.00	261,632
Cyber Frontier	$5.25	$15.75	1,048,576
OrcaCyber Zero 1.5	$7.50	$18.75	999,488

Full price list and tool fees · Every model · API docs

Questions

Per its own site in November 2026, an autonomous offensive security platform that tests web applications and APIs and reports findings it has validated with a working exploit.
Its pricing page listed no prices on 11 November 2026. It describes usage-based pricing scoped to your environment, sold through a demo, a quote or a cloud marketplace. Ask XBOW for a figure.
No. It is a model API. It answers requests from your agent or script, and the workflow, the tools and the approvals are yours.
Yes, that is the intended use, for authorized testing. Any agent that takes an OpenAI-compatible or Anthropic-compatible endpoint can use it, and the agent guides cover OpenCode, Claude Code, Cline, aider and others.
Its platform page says XBOW benchmarks models and routes each task itself. The pages we read in November 2026 did not describe bringing your own model.
Per token from a prepaid balance with a $10 minimum top-up by card and no contract. On 2026-11-12 input prices for security builds start at $0.40 per million tokens.

One key, from $10

PentestLLM is for systems you own or have written permission to test. The acceptable use policy says what is not allowed.

Get a key

usage, 1,000 agent turns on Qwen3.8 27B Cybersecurity

Input, 8,000 tokens at $0.40 per million: $0.0032
Output, 1,000 tokens at $2.40 per million: $0.0024
One turn: $0.0056
1,000 turns: $5.60

output price per million tokens, 9 builds

$1.3376 Gemma 4 26B A4B CybersecurityOrcaCyber Zero 1.5 $18.75