stylobot stylobot v8 Paid, self-hosted editions are now available. From £25/mo.
stylobot Logo stylobot

Stylo.Bot Live Reporting and Management Dashboard Now Available

Self-hosted · Realtime · Zero usage charges

Self-hosted behavioural bot protection that's simple to set up

Install StyloBot, point it at your site, and let it handle unwanted traffic. It runs in your own infrastructure, with no usage charges.

Live · Your Detection Bot
04:53:35 · 0ms

Top Bots

Name Bot % Threat Hits 1h Seen
GB zgrab 6199 {l:78,78,81,81,78,78,78,86,83,78,81,78,89,78,81,78,81,78,81,72} now
CA Cloud Provider 715 {l:0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,74,74} 43m
GB Headless Chrome 245 {l:0,22,0,33,0,0,53,0,22,17,17,22,0,22,11,0,0,0,0,44} 18h 0m
CA Unclassified 241 {l:0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,51,91,87,93} 15h 15m
NL Chrome 124 macOS 112 {l:0,0,32,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,39} 11h 23m
1–5 of 131
1
Not a demo. Your real request, live detection engine.
Non-blocking on stylo.bot. Every detector scores you and the dashboard records it. This demo throttles bots but blocks nothing, so you can watch the engine work on your own request. For production we recommend blocking confirmed bots. Each self-hosted install sets its own action per policy (allow, throttle, challenge, block).

Why StyloBot

Your traffic, your data, your system. It adapts on its own.

Your traffic, your data, your system

Self-hosted in your own environment. Visitor data never leaves it.

Adapts to your site on its own

It learns what normal traffic looks like for you. No rules to write, no tuning to schedule. Every control is still yours if you want it.

Zero usage charges

You pay for a licence, not per request. A traffic spike or a DDoS does not change your bill.

Free engine, paid management

The open-source build runs the full detection pipeline. Commercial adds fleet management, reporting and live config.

Protection that learns what normal traffic looks like.

StyloBot watches what a visitor actually does: request timing, page visits, browser consistency, and session behaviour. Small checks work together, behavior memory, signatures, reputation, and feedback loops combine into decisions that improve as traffic repeats. LLMs are optional side analysis, not the thing you depend on for every request. No SaaS. Nothing about your visitors leaves your environment.

How behavioral inference works →

Useful on day one. Start in observe mode and tune from real traffic
Every decision has a trace: raw signals, derived signals, detector deltas, aggregation, and policy action
Session behavioral analysis catches automation that rotates IPs, user agents, and fingerprints

Per-request output for this request, live

probability
100 %
confidence
100 %
risk band
VeryHigh
action
block

Plus detector breakdown, raw and derived signals, threat score, intent classification, policy action, and narrative reasoning.

Easy to start. Clear when you need answers.

StyloBot works with sensible defaults. When you want more control, you can see why it made a decision and change how it responds.

StyloBotDataDomeCloudflareHUMAN
Licence and usage modelFree FOSS / predictable commercial licenceSales-led SaaSEnterprise planCustom
Self-hostedYesNoNoNo
Open sourceFull engineNoNoNo
Usage chargesNeverOftenPlan-dependentCustom
Raw traffic sharingNot requiredCloud-scoredCloud edgeCloud-scored
Policy-controlled frictionYesYesYesYes

Choose your documentation path

Use the customer docs for setup, configuration, and rollout, and the GitHub repo for deep detector internals, architecture, and API references. Already use Compose? Start with the integration guide.

Live on stylo.bot right now

This is not a demo. It is the same detection engine checking every request to this site in real time.

Live Activity

Name Bot % Threat Hits 1h Seen ▼
GB zgrab 6199 {l:78,78,81,81,78,78,78,86,83,78,81,78,89,78,81,78,81,78,81,72} now
SG Mobile Safari 13 iOS 2 {l:0,0,0,0,0,0,0,0,0,0,0,0,0,33,0,0,0,0,0,33} 35s
US zgrab 19 {l:0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,58} 2m
KR Mobile Safari 13 iOS 3 {l:0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,33} 2m
SG Mobile Safari 13 iOS 2 {l:0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,33} 9m
SG PetalBot 9 {l:0,0,0,0,33,0,0,0,0,0,0,0,0,0,0,0,0,0,0,33} 11m
BR Mobile Safari 13 iOS 2 {l:0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,33} 13m
US Chrome 39 macOS 1 {l:0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,33} 22m
SG Mobile Safari 13 iOS 2 {l:0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,33,0,0,33} 22m
HK Mobile Safari 13 iOS 4 {l:0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,33} 23m
1–10 of 380
1

From the development blog

How StyloBot was built and the thinking behind it.

All articles →
Core concept
Behavioral inference systems
The architectural idea behind StyloBot: inferring intent from request behavior rather than matching patterns. Why that distinction matters for accuracy and for false-positive rates.
All bot detection posts
The full development series
30+ articles covering detector design, signal composition, LLM integration, HNSW session clustering, fingerprinting, and the commercial layer. Written as StyloBot was built.
Browse all articles →

Start simple. Grow when you need to.

The open-source edition is for starting StyloBot and leaving it to work. Paid plans add live control, detailed dashboards, shared access, reporting, fleet management, and support when you need to take a more active role. Your bill never rises because your traffic does. Try the commercial tools for 30 days with no card.

Open source and charity projects: contact us for a complimentary license.