Skip to content
Port 443 unlimited bandwidth

Dedicated Trojan Proxy — Port 443 TLS Routing on Datacenter or 23 Named ISPs

An encrypted TCP tunnel on a static IP, on the same port as ordinary HTTPS. Your client presents a domain-to-connect you can change from the dashboard, from $2.95 per IP per month with unlimited bandwidth.

trojan-client.json json
// all four values from your dashboard
{
  "server": "203.0.113.42",
  "server_port": 443,
  "password": "dashboard-password",
  "domain_to_connect": "boredgamings.xyz"
}
V2RayN · V2RayNG · V2Box TCP only
Protocols Trojan Shadowsocks SOCKS5 HTTP WireGuard
$2.95 per IP / mo
Unlimited bandwidth
3,830 dedicated IPs
23 named ISPs
99.99% uptime SLA
TCP only, port 443 Domain-to-connect you control 7-day instant cash back

Works with your stack

Python Node.js JavaScript TypeScript Go PHP Ruby Rust C++ Kotlin Swift Selenium Scrapy cURL Pytest Docker Kubernetes GitHub PostgreSQL Redis

Protocol

What Is a Trojan Proxy?

Trojan is an encrypted proxy protocol whose transport is built to look like ordinary HTTPS. The client opens a TLS connection to the endpoint on port 443 and the proxied stream travels inside it. To anything watching the network, that is a website.

The routing decision is made on the SNI — the hostname the client asks for during the TLS handshake. A request for your configured domain-to-connect is routed to the proxy; anything else falls through to a real web server. That is why the domain matters as much as the IP.

Dedicated means the address is assigned to your account, not drawn from a shared pool. ProxyPerk’s Trojan estate is approximately 3,830 static IPs across datacenter and ISP ranges, with unlimited bandwidth on every one.

What the camouflage is — and is not

It is a real TLS handshake on a real port, plus SNI routing that sends non-matching requests to a website. That shape is what makes the flow look unremarkable.

It is not an obfuscation plugin layer. There is no traffic shaping, no pluggable transport and no TLS-mimicking module bundled with this product. A network that classifies flows by behaviour and entropy can still classify an encrypted Trojan flow.

For a product built specifically to defeat deep packet inspection, see WireGuard Amnezia.

Boundaries

What Trojan Is Not

Four limits worth knowing before you buy, because each one changes which protocol you should pick.

No UDP

TCP only. QUIC and HTTP/3 will fail unless your client falls back to HTTP/2. For UDP, use WireGuard.

No obfuscation plugin

The camouflage is TLS shape plus SNI routing. It is not a DPI-defeating layer.

One IP, one identity

Static and exclusive. It does not rotate, and it is not shared with anyone else’s traffic.

No trial

Encrypted products are not trial-eligible. The 7-day instant cash back applies instead.

Pricing

One Base Price, Unlimited Bandwidth

Trojan starts at $2.95 per IP per month on datacenter. A named ISP carrier multiplies the base. Duration and quantity discounts stack on top of that — the calculator below does the arithmetic for your exact basket.

Bandwidth

Unlimited

no per-GB meter

Inventory

~3,830 IPs

datacenter + 23 ISPs

Guarantee

7 days

instant cash back

Estimate your bill

Base price × ISP multiplier, then duration and quantity discounts. This is an estimate, not a quote.

Estimated total
$0.00

Datacenter IPs are trial-eligible. Every product carries the 7-day instant cash back guarantee.

Start in the portal

Specifications

What You Get

Every endpoint is a managed service, not a bare IP. Here is what ships with it, and what does not.

Transport

TLS over TCP on port 443 — the same port as ordinary HTTPS.

Routing

SNI-based. Requests for your domain-to-connect are proxied; anything else reaches a real web server.

Credentials

Password, port and domain-to-connect shown in the dashboard. Dual auth with an IP allowlist.

Bandwidth

Unlimited on every static IP product. No caps, no overage.

Carrier choice

23 named ISPs plus datacenter, filterable by country, state and city.

Abuse log

A live 24-hour log with ban type, reason and escalation. IPs are replaceable.

trojan-check.sh bash
# credentials + domain from your dashboard
$ curl -x "socks5h://127.0.0.1:1080" \
    https://api.ipify.org
"203.0.113.42"

# the endpoint answers TLS on 443
$ curl -sv https://"boredgamings.xyz" \
    --resolve boredgamings.xyz:443:203.0.113.42 2>&1 | head
* TLSv1.3 connection established
V2RayN · V2RayNG · V2Box TCP only

What the dashboard shows

  account-levelper IP
Domain-to-connect one value—
Password one valueone value
Server + port —per IP
TCP fingerprint one valueone value

The domain-to-connect is one setting for the account. Change it once and every Trojan client on that account picks up the new hostname.

Full reference: API documentation · Client walkthroughs

Short answer

A dedicated Trojan proxy is an encrypted TCP tunnel on port 443 with a static IP, from $2.95 per IP per month with unlimited bandwidth. The client presents a domain-to-connect as the TLS SNI, and that domain is one account-level setting you can change from the dashboard. The pool is about 3,830 datacenter and ISP addresses, with 23 named carriers selectable at purchase.

The honest part: the tunnel is TCP only, and there is no obfuscation plugin layer — the camouflage is the TLS shape plus SNI routing, so a network that classifies flows by behaviour can still classify this one. If you need to defeat DPI, use WireGuard Amnezia. Trojan is not trial-eligible; every purchase carries the 7-day instant cash back.

At a glance

Port
443 / TCP
Routing
SNI
Trial
Not eligible
Guarantee
7 days

Configuration

The Domain-to-Connect, Explained

The domain-to-connect is the hostname your client sends in the TLS SNI. It is the value the server matches on to decide whether a connection is proxy traffic or a visitor who found the address.

ProxyPerk shows it in the Proxy Credentials card next to the password, and you can change it from the dashboard at any time. It is one setting for the account — not one per IP. Change it once and every Trojan client on that account uses the new hostname.

That is the practical reason it exists: if a domain picks up a reputation problem, you rotate the hostname instead of the address, and your IPs stay where they are.

Why the SNI matters

A TLS client sends the hostname it wants before the certificate arrives. Trojan uses that to route: your domain goes to the proxy, everything else falls through to a real web server on the same address. Point the client at the raw IP instead of the domain and the SNI is wrong, which looks like a timeout rather than an error.

Check the match first

When a Trojan tunnel stops passing traffic, compare the domain in the client against the one in the dashboard. A mismatch is the most common cause, and it produces no useful error — just a connection that opens and then goes nowhere.

Step-by-step client walkthroughs: setup guides.

Setup

V2RayN, V2RayNG and V2Box

Four values from the dashboard — server, port, password, domain — and you are running. The local client opens a SOCKS5 listener that your application talks to.

The four values

Every Trojan IP in your dashboard shows its server address, port, password and the account-level domain-to-connect. Copy them into the client. There is no key exchange and no certificate to install.

Point at the domain, not the IP

Set the client’s server field to the domain-to-connect, not the raw IP. The SNI is what routes the connection, so a raw IP in that field produces a handshake that opens and then stalls.

How your app connects

The client listens locally on 127.0.0.1:1080 and exposes a SOCKS5 proxy. Point your application at that listener. Use socks5h:// so DNS resolves on the far side of the tunnel.

tunnel.py python
# client already running on 127.0.0.1:1080
# socks5h = DNS resolves remotely
import requests

P = {"http": "socks5h://127.0.0.1:1080",
     "https": "socks5h://127.0.0.1:1080"}

r = requests.get("https://example.com/",
                 proxies=P, timeout=30)
print(r.status_code)
pip install requests[socks] TCP only

When it times out

1. Domain in the client matches the dashboard value.

2. Client points at the domain, not the raw IP.

3. Outbound TCP on port 443 is allowed locally.

Walkthroughs live in the setup guides.

Choosing

When Another Protocol Fits Better

Trojan is one of eight protocols in the lineup. It is the right pick when you want TLS-looking traffic on a static, dedicated IP with no bandwidth meter — and the wrong pick in these cases.

Your requirement Better fit
Plain static proxy, no client software in the stack SOCKS5 dedicated
An obfuscation layer for networks running deep packet inspection WireGuard Amnezia
A plain WireGuard tunnel with UDP support WireGuard
Rotating exits across many countries at a per-GB rate Rotating residential and mobile pools

Comparing the three proxy protocols in detail? Read Shadowsocks vs Trojan vs SOCKS5 — this page does not repeat it.

Guarantee

Test It on the Network That Blocks You

A spec sheet cannot tell you whether your own network will let a TLS tunnel on 443 through, and this product has no obfuscation layer to help it. Buy the smallest batch that covers one target, point a real request at it, and measure the result yourself.

Step 1

Buy a small batch — one IP is enough to test one target.

Step 2

Start the local client and point a real request at it from the network you need it for.

Step 3

Not delivered or not as described? Open a dashboard ticket within 7 days.

Every ProxyPerk product carries a 7-day instant cash back guarantee. Trojan is an encrypted product, so it is not trial-eligible — datacenter IPs elsewhere in the lineup are. The cash back is what applies here instead, requested through a support ticket and validated in real time against the published terms.

Covers non-delivery and service issues. Third-party software incompatibility and prohibited uses are excluded — Terms of Refund.

Help center

Trojan questions

The setup failures and policy questions people actually hit.