Days of setup, done in one apply
The Meet server, videobridges and Jibri recorders, with their networking, all declared in Terraform. Run it against your own AWS account and the whole stack comes up together.
Product
Ready-made Terraform scripts that build a complete Jitsi Meet platform in your own AWS account. Pick one by concurrent users, from a single 50-user server to an auto scaling, multi-shard setup for 2,000.
A single Jitsi server by hand is an afternoon. A cluster with separate videobridges, Jibri recorders and auto scaling is not, and doing it twice the same way is harder still. Infrastructure as code turns that into one command you can rerun.
The Meet server, videobridges and Jibri recorders, with their networking, all declared in Terraform. Run it against your own AWS account and the whole stack comes up together.
No missed firewall rule, no half-configured Jicofo. A second environment for staging is the same script with different variables, not a week of repeating steps from memory.
You get the code and a readme, not a subscription. When your setup grows, you change the code and apply again, instead of rebuilding servers by hand.
Up to 500 users, one shard does the job: one signalling server and its videobridges. Past that, the scripts split load across several shards behind HAProxy, so no single Jicofo or Prosody becomes the bottleneck.
50 to 500 concurrent users
1000 to 2000 concurrent users
Sized by concurrent users, from 50 to 2,000. Each one is delivered with a readme for your own AWS account. Contact us for pricing and to get the one that fits.
50 concurrent users
50 concurrent users
100 concurrent users
100 concurrent users
250 concurrent users
250 concurrent users
500 concurrent users
500 concurrent users
1000 concurrent users
1000 concurrent users
1500 concurrent users
1500 concurrent users
2000 concurrent users
2000 concurrent users
From 500 users up, the videobridges sit in an AWS Auto Scaling group. When calls get busy, new bridges start and register with Jicofo; when things quiet down, the extra bridges are removed. You pay for the peak only while the peak lasts.
One thing to check before you apply a large script: new AWS accounts often have a low vCPU quota, and a 2,000-user cluster will hit it. Raise the quota first, because approval can take a day or two.
Kubernetes works for Jitsi, but videobridges need their UDP port reachable from the internet, which fights the way Kubernetes networking usually works. For most teams on AWS, plain EC2 with Terraform is simpler to run. If you are set on Kubernetes, our Jitsi Kubernetes guide covers scaling bridges with Octo.
Size for your busiest meeting hour, not your average. If you would rather not run the scripts yourself, we can set Jitsi up on AWS for you. And if you only need a standard server, the Jitsi Meet marketplace images launch in a few clicks.
Terraform code that builds the Jitsi Meet server, its videobridges and, if you pick a recording script, Jibri servers in your AWS account. It comes with a readme for running it.
No. They are written for AWS only. For Google Cloud, our Jitsi Meet marketplace images are the fastest route. For Azure, we can set Jitsi up for you as a service.
Not for sale any more. Jitsi runs on Kubernetes, and our guide to scaling videobridges with Octo on Kubernetes shows how, but we only sell Terraform scripts for AWS today.
From 500 users, videobridges run in an AWS Auto Scaling group behind one signalling server. Bridges are added as load rises and removed when it drops. The 2,000-user scripts add HAProxy and Octo across shards.
Contact us with your concurrent user count and whether you need recording. We confirm the right script, agree the price and send the code with its readme. There is no online checkout.