What a credit is worth when the page never says
Six channels here meter in credits without printing what a credit costs. On two of them the figure divides out of published plan prices and allowances. On the other four nothing on the page reaches it, for four different reasons. As of 2026-09-22.
| Channel | What the page publishes | How close that gets |
|---|---|---|
| Hedra | Three plan prices against three monthly allowances | Divides out at about a cent, falling to half that |
| Synthesia | Credits a month, stated as minutes of video, with plan prices | Divides to a cost per minute, assuming full use |
| Luma | Credits per second by model and resolution | Stops one published figure short |
| Magnific | Credits against named seconds at named resolutions | Ranks its own models and converts to nothing |
| HeyGen | Plan prices and allowances, with consumption stated to vary | Cannot be reached: no quantity of output is fixed |
| Pika | Credits by feature, with no duration attached | Cannot be reached: two halves are missing |
Inclusion rule. Channels metering in credits whose pages do not state a credit price. Arithmetic shown here divides two figures published on the same page and adds nothing else. Order. Channels where the figure divides out first, then the ones where it does not.
1A division is a floor rather than a price
Dividing a plan by its allowance assumes the whole allowance gets spent and that nothing expires unused. Where the lifetime of a credit is unpublished, and on one of these pages it is raised in the FAQ and left open, the division produces the best case and reads like the normal one.
The gap is not small. A buyer using half of a monthly allowance is paying twice the divided figure, which is more than the spread between the cheapest and dearest plan on the same page. That is why the ledger keeps the blank.
2Four different reasons a credit cannot be valued
One page is missing a single sentence: it publishes credits per second and no credit price. One publishes exact quantities per model and no money. One states outright that consumption varies by model, duration and complexity, so no quantity of output is fixed at all. One attaches credits to features rather than to any duration.
Those are increasing distances from a rate. The first is a publishing decision, the last is a pricing philosophy, and only the first would be closed by the seller writing down a number it already knows.
3What a buyer can still do with an unpriced credit
Where credit figures are attached to named durations on named models, they rank a channel's own catalogue precisely, which is the decision a buyer already committed to that channel actually faces. The comparison that fails is the cross-channel one.
For anyone still choosing, the practical move is to treat a missing credit price as a reason to shortlist elsewhere and revisit if it appears. The three channels here that do publish both halves did it with one sentence each.
- Plan prices and monthly allowances$15 a month for 1,500 credits, $30 a month for 5,400 and $75 a month for 14,400
- The credit-to-minutes equivalence published per plan1,200 credits a month stated as 10 mins of video, and 3,600 credits a month stated as 30 mins
- Plan prices$29 a month on Starter and $89 a month on Creator, or $18 and $64 a month billed yearly
- Why no credit count per video is printedCredits are described as the single currency for video and asset creation, and the number a video uses can vary by model, duration and complexity
- What Magnific publishes instead of a rateCredits against named durations of video, with no price for a credit
- What Luma publishes instead of a rateCredits per second, with no credit price
- What Pika publishes instead of a rateCredits by feature, with no credit price
4What the seconds buy
Price aside, what goes into the seconds behind “Credits with no price” is described in SceneMixer's guides:
A short drama episode has two jobs: stop the scroll in its first seconds, and make the next episode feel necessary in its last.
SceneMixer, Writing Hooks and Cliffhangers Into a Short Drama Script (read 2026-10-13)
SceneMixer turns a novel into a watchable short-drama series in one pipeline: upload the novel (Chinese or English) → AI extracts characters, scenes, and an episode-by-episode structure → generates consistent character and scene art → writes storyboard scripts → renders video segment by segment (16:9 by default, 9:16 in one setting) → composites full episodes.
SceneMixer, Turn a Novel into a Video Series with AI (read 2026-10-13)
5Sources
Each figure is read from the vendor page linked in the ledger, checked 2026-09-22. How a figure qualifies is on method; what a rate leaves out is on limits and unspent credits and parallel generations. Nearby: Who built the model, Fixing a clip, Draft then finish.