The complete Hyx feature inventory.

Visitors use the fastest connection their browser supports for a quick and reliable experience.

Returning visitors can reconnect in less than one network round trip.

A single edge location served 1.13 million requests per second in testing.

Requests are handled independently so one slow operation does not stall unrelated traffic.

Popular pages and files are served without contacting your server again.

We follow your site's caching rules for what to store, how long to store it, and when to refresh it, so visitors receive current content quickly.

Update a page and purge the old version from every location worldwide immediately.

Pages and files are compressed ahead of time and delivered in the most compact format each browser accepts, without adding request-time work.

Images are converted to lighter modern formats for each visitor. The dashboard reports the resulting bandwidth savings.

Every connection is secured with modern encryption that's been tuned to be as fast as it is safe.

Your security certificates live in high-speed memory and update instantly, so connections are secured without ever pausing to look them up.

We keep the proof that your certificate is valid fresh and attached automatically, so browsers never stop to double-check.

On the first connection, Hyx learns how your server prefers to communicate and reuses that information on later requests.

If a request to your server fails, Hyx retries it only when doing so is safe.

Hyx avoids repeating work or waiting on extra services while handling a request.

Production traffic profiles guide optimization of common request paths.

Each location adjusts its capacity to the available hardware.

Hyx upgrades beneath live traffic without dropping active connections.

Traffic controls protect Hyx and your server when demand exceeds capacity.

System components are isolated so a fault in one area does not take down your site or disable protection.

Requests and responses follow HTTP framing and forwarding rules across the full proxy path.

Firewall rules are prepared before use so each request can be checked quickly.

Write rules that match the URL, country, device, headers, request size, or a visitor's recent behavior.

Matching rules can block, challenge, rate-limit, bypass cache, tag a visitor, or suppress a known false positive.

Managed industry rules check each request for common web attacks.

The full rule set checks each request in about two millionths of a second.

Turn protection on with a switch, and quiet any single rule that's too noisy for your particular site.

Reputation data identifies known-malicious addresses before HTTP request processing.

Decide which countries can reach your site, with checks so fast they stay instant even in the middle of an attack.

Limit request rates per visitor to control abuse while allowing normal customer traffic.

Ban an address a single time and it's turned away at every location worldwide, before it can even connect.

Suspicious clients receive a browser verification challenge before access is allowed.

Search engines like Google are recognized and never blocked, challenged, or mistaken for attackers.

Enable recommended browser protections, including enforced HTTPS, per site without changing application code.

We watch for waves of failed logins and catch password-guessing attacks even when they're carefully paced to look normal.

Visitors that repeatedly attempt attacks remain flagged until they complete a human verification challenge.

Lightweight checks record suspicious behavior for later firewall and model decisions.

At connection time, Hyx captures a signature of the client's software that you can filter or use in rules.

Connection behavior can identify common bots before they send an HTTP request.

Trusted browser signatures cut down on false alarms, while known-bot signatures unmask bots pretending to be real browsers.

Connection identity works across the browser connection methods supported by Hyx.

Fine details of how a browser communicates add another signal to how confidently we can tell humans from bots.

Requests are tagged as browser or automated traffic for dashboard filtering.

Hyx AI reviews each request on arrival so policies can stop attacks before they reach your site.

Hyx AI reviews each visitor using recent behavior and request details.

High-confidence attacks can be stopped on the first request without a warm-up period.

Hyx AI uses recent visitor behavior to identify abuse patterns.

Calibrated scores keep a 90% result consistent enough to support clear policy thresholds.

Hyx AI learns from real traffic and releases approved versions across the network automatically.

Hyx AI updates must pass false-positive and detection checks before release.

Training uses high-confidence labels from traps and completed challenges to reduce ambiguous examples.

Confirmed mistakes become lessons for future Hyx AI versions.

Low-confidence cases are queued for human review and can become labeled training examples.

Block records include the signals and policy behind the decision.

Run Hyx AI in watch-only, challenge, block, or automatic mode. Automatic mode blocks high-confidence bots and challenges borderline cases.

Clients repeatedly blocked by Hyx AI can be promoted to a persistent ban.

Quality checks track Hyx AI accuracy and consistency after release.

Roll every location back to the previous trusted version with one switch, or let automatic safeguards reduce enforcement.

A built-in assistant suggests rules for emerging attacks. Quality checks and human approval are required before publication.

Candidate Hyx AI versions run beside the live version without affecting protection decisions.

New models can be evaluated at one location before network-wide rollout.

Network-level aggregation can detect related attacks distributed across many addresses.

Request logs include connection details, delivery status, Hyx AI decisions, and serving time.

Block records show the trigger, Hyx AI version, and location that handled the request.

Explore traffic history from the dashboard with a safe point-and-click query builder.

Built-in detectors label suspicious behavior such as SQL injection, Log4j probes, and login attacks.

The dashboard highlights frequently triggered rules with low confirmed value for review and tuning.

Share dashboard views of traffic, security, bots, origin health, and day-over-day trends with your team.

Origin health views highlight slow responses, errors, and timeouts so teams can find affected servers quickly.

Ask questions in plain English and receive answers drawn from dashboards scoped to your sites.

Review Hyx AI status, uncertain cases, accuracy, and rollback controls from one screen.

Operational views expose certificate renewals, data retention jobs, and other background tasks.

Manage domains, origins, certificates, and rules from one place, with changes distributed across the network.

Add a domain and Hyx issues and renews its security certificate automatically.

Rules, certificates, blocklists, and switches propagate across the network within seconds without a redeploy.

Global settings apply at every location, with per-site overrides where needed.

Each network location runs the same software and sizes itself to the available hardware for consistent operation.

Request records include the serving location and active protection configuration.

Traffic history and Hyx AI versions are retained for audits and rollback.

Manage backups, files, malware scans, plugins, themes, performance settings, logs, and billing alongside the CDN.