Skip to content
FullsendAI
Log inGet started

AI for security research

AI for security
research.

Nef 4.0 and Sifu 3.5 are our own models, built for offensive and defensive security research. Use them in the chat or your own tools, hosted in the USA.

Try Nef 4.0Get API Key

Built for the work.

From the first recon notes to the detection that closes the gap, on systems you’re authorized to test.

  • Penetration testing

    Plan an authorized engagement, triage scan output and work through findings with a model that keeps up.

  • Exploit analysis

    Read a public PoC or advisory, understand the root cause and judge real-world impact on your own systems.

  • Security code review

    Audit diffs and whole repositories for injection, auth and memory-safety bugs, with fixes you can apply.

  • CTFs

    Reverse a binary, decode a protocol or work a web challenge, step by step, for practice and competitions.

  • Malware triage

    Make sense of samples in an isolated lab: deobfuscate scripts, read decompiled code, extract indicators.

  • Detection engineering

    Turn findings into Sigma, YARA and SIEM queries, and test them against the behaviour you saw.

Our flagship models.

Nef 4.0 and Sifu 3.5 are built for authorized offensive security: red team and pentest work, security research and code. Made for hackers, coders and security researchers, and just as good for everyday work.

FlagshipNot rated

Nef 4.0

Our strongest model, built for authorized offensive security: red team and penetration testing work other models refuse, plus deep code review and research.

  • Red team and penetration testing
  • Exploit development and validation
  • Hard, multi-file coding
$0.60 in · $2.40 out per 1M tokensTry Nef 4.0 in the chat
Fast and affordableExcellent

Sifu 3.5

Quick and low-cost, for engagement triage, security research and code review at volume, and for everyday work.

  • Engagement and recon triage
  • Security research at volume
  • Everyday coding and writing
$0.30 in · $1 out per 1M tokensTry Sifu 3.5 in the chat

For authorized testing and security research, as our Terms of Service require.

Two ways to work.

The same models and the same balance in the chat and your own tools.

In the chat

Open Nef 4.0 in the web chat. Paste code, logs or a crash trace, attach files, and keep every conversation in your account.

Open the chat

In your coding agent

Run npx fullsendai connect: it signs you in and sets up Claude Code, Codex, Cursor, Cline and more with Nef, Sifu and every other model we serve. Or call the OpenAI- and Anthropic-compatible API from your own scripts.

Connect your agent

Your terminal, your tools.

Point Claude Code, or any OpenAI-compatible tool, at FullsendAI and keep your Kali or lab workflow. Every call is billed per token from one balance.

export ANTHROPIC_BASE_URL=https://fullsendai.com
export ANTHROPIC_AUTH_TOKEN=fs_live_...
export ANTHROPIC_MODEL=nef-4.0
export ANTHROPIC_DEFAULT_HAIKU_MODEL=nef-4.0
claude

Authorized work only.

Nef 4.0 and Sifu 3.5 are for security work you have permission to do.

Test systems you own or have written permission to test, analyse samples in an isolated lab, audit code, and build detections and defences. Attacking, accessing or disrupting systems without authorization, or creating malicious software to use against others, breaks our Terms, and we may suspend accounts that do.

Read the acceptable use policyHow we handle your data

Questions, answered.

What security teams ask before the first engagement.

FullsendAI’s own models, built for authorized offensive security, security research and code. Nef 4.0 is the strongest; Sifu 3.5 is fast and low-cost, for work at volume.

Authorized work: testing systems you own or have written permission to test, analysing malware in an isolated lab, auditing code, and building detections and defences. Attacking systems without authorization is against our Terms of Service.

API request and response bodies aren’t stored unless you turn on body storage for a project, and then only for 30 days. Web chat conversations are saved to your account until you delete them. Nothing is used for training.

Yes. FullsendAI speaks the OpenAI and Anthropic APIs, so Claude Code, Cursor, Cline, Aider and the SDKs connect with a base URL and one key. That key opens every other model we serve too, so you can switch from your tool's model picker.

Start your next engagement.

Get a key in a minute and pay only for the tokens you use. Every model hosted in the USA.

Get API Key