Take their largest published plan, Scale, and its 24 TB of annual bandwidth. Two people watching a 720p channel around the clock for twelve months deliver 23,652 GB. That is the whole year's allowance spent on an audience of two, and the third continuous viewer costs another $3,548 at their published rate. Bandwidth pricing assumes a customer who uploads a library, publishes an event, and pays for the traffic it draws. A linear channel is a permanent consumer of bandwidth by design, so a plan sized in terabytes per year is being asked to do the one job its unit was not chosen for.
The published price here is $254 a month billed yearly, per channel, and delivery is unlimited from 300+ cities on every plan. The night the channel goes viral costs what the quietest Tuesday costs. What a TV channel costs runs both models out over a full year.
Somebody has to run the encoder
The cost that never appears on a pricing page is the machine. A passthrough platform takes an RTMP or HLS push, which means the schedule lives on hardware you own: a PC in the office running encoder software every hour of every day, on your power, your internet connection and your uptime. It runs at 4am on a public holiday or the channel is dark. Anyone who has kept one alive for a year knows the failure modes, and most of them happen while nobody is watching the machine.
Our engine holds the schedule instead. You upload the library, build the week in the browser, and the channel plays from the platform with gap auto-fill and instant restart behind it. Dacast's published material describes VOD to Live, which starts a single pre-recorded video as a live event at a chosen time, and playlist channels, which play a list inside the player with autoplay, optional looping and viewer skipping, so two viewers can be watching different things at the same minute. A station clock that every viewer shares, running unattended for months, is a different mechanism and is not described in those materials.
Where the breaks go
The same split shows up in advertising. Dacast supports VAST, VPAID and VMAP tags entered in their dashboard and called by the player on the viewer's device, so the ad exists only where a player script runs. A linear channel monetizes differently: breaks sit on a continuous timeline, each one has to declare what kind of break it is and who is entitled to fill it, and the ad has to be stitched into the stream so it arrives identically on a browser, a phone, a smart TV and a set-top box that never runs a player script. That signalling is what SCTE-35 does and what server-side insertion does with it, on every plan, with your own demand and 100% of the revenue. Cue signalling, guide feeds and as-run logs are not described in Dacast's published material, so this page says that and nothing stronger.
What sizing by viewers actually buys you
Plans here are sized by concurrent viewers rather than gigabytes: 100 on Launch, 500 on FAST, 1,000 on Studio, and roughly double each again with ads enabled. Every one of those numbers is printed on the pricing page, so picking the right plan is a decision you make once, in advance, with a calculator and without booking a call. Inside a tier the bill does not move at all. A channel can have its best month in three years and the invoice reads exactly the same as the month before, which is the one thing a bandwidth meter cannot offer at any price.
Testing it costs a week
Other platforms sized for events and libraries price the same way, and the Muvi comparison runs the arithmetic against streaming hours rather than gigabytes. There is no contract here, plans cancel any month, and the first purchase carries a 7-day money-back guarantee. Put one channel on air here for a month, keep whatever you are running now, and compare the two invoices with the bandwidth line highlighted. Tell us what you are launching.