The price you see before converting a clip is not a flat per-video fee and it is not a subscription tax. It is computed from three things: how long the clip is, what resolution it is, and whether you are on a plan. Here is the exact math.
The two resolution tiers
Cost scales with how much the engine has to process, and that scales with pixel count, so there are two tiers:
- HD and below (up to 1920×1080): $0.60 per second.
- Above 1080p (anything wider than 1920 or taller than 1080): $1.275 per second.
A 10-second 1080p clip is $6.00. The same 10 seconds at 4K is $12.75. Every conversion has a $0.60 minimum, so a 1-second test clip still costs 60 cents, not a fraction of one.
Where the rate comes from
Conversion is metered work, an AI model processes every frame, and longer or higher-resolution clips take proportionally more of it. Charging per second and per resolution tier passes that scaling straight through instead of averaging it into a flat fee that overcharges short clips and undercharges long ones.
How to estimate your cost before uploading
The math is simple once you know two numbers: your clip's exact duration in seconds, and which resolution tier it falls into. Multiply duration by the per-second rate for that tier and your plan, round up to the nearest cent, and apply the $0.60 floor if the result comes in under it. You never actually have to do this by hand, the tool reads duration and resolution straight out of the file in your browser and shows the exact quote before you spend anything, but knowing the math means no surprises.
Two details catch people off guard:
Billing is by the second, not the whole clip. A 12.4-second 1080p clip is not rounded up to 13 seconds, it is billed for 12.4 seconds exactly, $7.44 at the base rate. Only the final price rounds up to the nearest cent.
The resolution check is exact pixels, not the label on your export settings. A clip only qualifies as "HD and below" if it is 1920 pixels or narrower on its long edge and 1080 pixels or shorter on its short edge, checked without regard to orientation, so a 1080×1920 vertical export prices exactly the same as a 1920×1080 landscape one. But some cameras and encoders pad a frame's height to a multiple of 16 for encoding reasons, landing at 1088 pixels instead of a clean 1080 on the short edge. That clip is still commonly labeled "1080p" everywhere you look at it, but 1088 is more than 1080, so it crosses into the higher tier and roughly doubles the per-second rate. If a quote looks higher than you expected for what you thought was a 1080p file, check the file's actual pixel dimensions, not its nominal resolution.
What lowers the rate
The per-second numbers above are the base rate, what anyone pays with no plan. Subscribing lowers the multiplier applied to the same wholesale cost:
- No plan: base rate, as above.
- HDR-20 ($20/mo): a modest discount.
- HDR-50 ($50/mo): a larger discount, down to $0.552 per second at 1080p and below, $1.173 above.
- HDR-200 ($200/mo): the lowest per-second rate, $0.48 per second at 1080p and below, $1.02 above.
A subscription is a monthly balance, not a per-video coupon: you pay the plan price, that dollar amount is credited to your balance every month, and every conversion you run draws from it at the plan's lower rate until it is spent. Unused balance from a subscription does not roll over month to month; a separate pay-as-you-go top-up (any amount from $50) never expires and is spent only after your subscription balance for the month is used up.
A worked example: eight highlight clips a month
Take a wedding videographer delivering eight 45-second highlight reels a month, each exported at 1080p for wedding clients to share. That is 360 seconds of conversion a month. Here is what each pricing path actually costs:
| No plan | HDR-50 | HDR-200 | |
|---|---|---|---|
| Monthly subscription | $0 | $50 | $200 |
| Rate at 1080p | $0.60/sec | $0.552/sec | $0.48/sec |
| Cost per 45-second clip | $27.00 | $24.84 | $21.60 |
| Cost for 8 clips | $216.00 | $198.72 | $172.80 |
| Actual monthly spend | $216.00, all top-up | $198.72 total: $50 subscription + $148.72 top-up | $200.00, subscription covers it, $27.20 unused |
Three things fall out of that table that are not obvious from the per-second numbers alone:
HDR-50 is the better deal here even though HDR-200 has the lower rate. At this volume, HDR-200's $200 balance is bigger than the $172.80 of work actually being done, so $27.20 of it goes unused every month and does not carry over. HDR-50's smaller balance gets fully spent and topped off at its own discounted rate, landing at $198.72 total, cheaper than either alternative.
HDR-200 only pulls ahead once usage actually fills the balance. The break-even point is $200 of usage at the $0.48 rate, just over 417 seconds a month, a little more than nine 45-second clips. Below that, some of what you are paying for goes unspent. Above it, every additional second is at the lowest rate the platform offers, and HDR-200 wins outright.
HDR-20 is sized for occasional use, not eight clips a month. Its $20 balance covers at least 33 seconds of 1080p footage even at the plain base rate, closer to one short clip every couple of weeks than a standing production workflow. It exists for the person who converts something now and then and would rather not think about a top-up every time, not for volume.
None of this changes what the conversion itself does, the file coming back is identical regardless of which balance paid for it. It only changes what you pay to get there.
Volume also has to be consistent for a plan to pay off, not just high in a good month. Drop the same videographer to four clips in a slow month, 180 seconds, and the numbers shift: base rate comes to $108.00, HDR-50 comes to $99.36 total ($50 subscription plus $49.36 top-up), and HDR-200 comes to a flat $200 for $86.40 of actual work, the worst outcome of the three that month. A plan sized for your busiest month, not your average one, quietly overpays every slower month in between.
Why per-second beats a flat fee
Metered, per-second pricing looks more complicated than a flat per-video price, but a flat fee only looks simpler until you check what it actually charges for different footage.
Picture a flat $10-per-video price instead, just to see what it would distort. A 3-second 1080p test clip actually costs $1.80 to convert at the base rate, so a flat $10 charge would be more than five times the real cost for someone just checking whether their footage is worth converting at all. A 55-second clip above 1080p actually costs $70.13 to convert, so the same flat $10 would be selling seventy dollars of processing for ten, a loss that either has to be priced into everyone else's rate or absorbed until it is not sustainable.
A flat fee always does one of those two things: overcharge the short, cheap-to-process clips to subsidize the long, expensive ones, or undercharge the long ones and eat the difference. Per-second, per-resolution pricing does not average anything away. A 3-second test costs what a 3-second test costs. A 55-second 4K delivery costs what that actually takes to process. Neither one is subsidizing the other, and the quote you see before you commit is the same number that gets charged, not an estimate that gets reconciled later.
What you are not charged for
- Failed conversions are refunded automatically, usually within minutes, no request needed.
- Uploading and previewing cost nothing. You see the exact quote, resolution and duration read directly from the file in your browser, before you spend anything.
- There is no separate fee for HDR10 metadata, no watermark removal fee, no download fee. The quoted price is the full price.
Then what
For what you are actually paying for, mechanically, see how to convert an SDR video to HDR. For why some clips are worth converting and others are not, regardless of price, see SDR vs HDR.