You can see every layer.
Because you own every layer.
No black box, no vendor dashboard you rent by the message. NGINX, PHP, HyperNode, Redis, MySQL, Kannel, the SMSC: the real stack, running on your hardware, open to your inspection. When something matters at 3 a.m., you can actually look.
No black box · air-gap ready · you hold the console
Why ownership isn't just a feature
When it matters at 3 a.m., do you file a ticket, or open a shell?
Nothing leaves your walls
The system of record, the queues, the archive, the search: all on your hardware. Air-gap it if you need to. No message ever depends on someone else's uptime.
Inspect anything
SMPP on :2775, the HTTP API, PM2 workers under watch. It's your server: open a shell, read the logs, prove it to yourself.
No per-message tax
The pipeline is yours. Send a thousand or a billion: the architecture doesn't meter you, and neither does anyone else.
The whole stack, named honestly
Every box is real. Every box is on your server.
Nothing hidden, nothing rented. Name any layer, update it, inspect it: it's a component you hold, not a service you rent.
Built on tools you already trust
No proprietary black box. Familiar parts, all the way down.
Ubuntu
Red Hat
Bootstrap
HTML5
JavaScript
Node.js
PHP
C++
MongoDB
MySQL
Redis
Elasticsearch
NGINX
HAProxy
PM2
Git
Docker
KubernetesEvery part is a named, standard component you can look up.
Yours, down to the last word
Run it in your customers' language.
A built-in translation manager puts every phrase in the platform in your hands: search it, edit it, improve a wording, or add a whole new language. No re-deploy, no vendor ticket. The interface your resellers and their clients see speaks the way they do.
- Thousands of phrases, editable in place
- Track missing and review-needed strings per locale
- Add languages yourself, RTL included
Hold the console
It's your server. Prove it to yourself.
Open a shell and look. The SMPP server answers on :2775, the HTTP API is right there, and every worker is under watch. No dashboard stands between you and the truth of your own system.
$ pm2 list
queue-workeronline
schedule-workeronline
routing-workeronline
mnp-lookuponline
archive-workeronline
$
Deploy it your way
From a single box to active-active.
You can't promise high availability on one box. So when you sell uptime to your own clients, Active-Active is a real, tested tier (both nodes live, sharing load), not a special project.
Basic
Single node. Everything on one box, perfect to start.
Active-Standby
Warm failover. A second node waits, ready to take over.
Active-Active
Both nodes live, sharing load. For operators who sell uptime SLAs to their own clients.
Custom
Sized with you, for the deployment your scale actually needs.
Workers, watched
PM2 and Supervisor keep every worker alive; monitors wire new ones in. Nothing runs unwatched.
Numbers, stated honestly
Throughput, latency and uptime are published only once measured under stated conditions. No invented figures.
Air-gap ready
Every layer, including the native AI, runs on your servers. Behind a DMZ, or fully off the grid.
One server, rented from anyone, sized to your traffic.
The question every buyer asks first, answered before you ask. These are the planning figures from our architecture guide, not contractual minimums: message size, delivery-receipt volume and reporting load move them.
| Sustained traffic | Monthly volume | Starting point | First thing to watch |
|---|---|---|---|
| ~1 msg/s | Up to ~1,000,000 SMS | 4 vCPU, 8 GB RAM, 100 GB SSD, everything on one box | Disk filling with logs and delivery receipts |
| ~10 msg/s | ~1,000,000 to 10,000,000 | 8 vCPU, 16 to 32 GB RAM, 250 to 500 GB SSD | Search and reporting under client load |
| ~50 msg/s and up | 10,000,000 and up | 16+ vCPU, 32 to 64 GB RAM, fast NVMe, components split across nodes | Single-node limits: time for Active-Active |
Planning guidance from the Smppcube architecture guide. Peaks, delivery receipts and retention stress a server far more than the monthly total.
Starting from zero, but need 100 to 200 msg/s of headroom?
A dedicated 4 to 8 vCPU, 32 GB server covers it while your monthly volume is still low, and typically costs $50 to $100 a month. You move up a tier when sustained volume, delivery receipts and reporting say so, not before.
Where to rent it
Any provider works if the CPU is dedicated, not shared. Hetzner and Netcup in Europe (we run on Netcup ourselves), DigitalOcean and OVHcloud worldwide, Alibaba Cloud for a data centre inside Saudi Arabia. Your own hardware behind your firewall is equally welcome, air-gapped if you need it.
- Hetzner
- Netcup
- DigitalOcean
- OVHcloud
- Alibaba Cloud
- or your own hardware
Three lines. Nothing hidden.
$6,400 once: the license, every channel, every feature and year one of support.
$50 to $100 a month: the server, paid to the provider you choose, in your own account.
Your carrier traffic, at the rates you negotiate with your own SMSC or aggregator. We never sit in the middle.
Optional, and only if you want them: support renewals after year one, extra production installs, training days. Nothing per message and nothing per seat, ever.
Before you ask
Can I run this fully air-gapped?
Yes. Every layer, including the native AI, runs on your servers. Behind a DMZ or fully off the grid, the platform keeps working.
Do I get root on my own box?
It's your server. You install it, you hold the keys, you can inspect every process, log and queue. That's the point of owning it.
What about high availability?
Active-Active is a real, tested deployment tier where both nodes carry load. Choose it if you sell uptime to your own clients: you can't promise high availability on a single box.
Own the stack. Every layer. Every log. Every key.
The whole platform, on hardware you control: no black box, no per-message tax, nothing rented.