
- Credits per clip
- 300
- Total credits
- 600
- Clips covered by one 600-credit pack
- 2
- 600-credit packs needed
- 1
A failed generation does not consume credits, but a retry is another full-cost job: budget the retries you expect, not the clips you hope for.
Creative decisions feel free and cost credits. This is the arithmetic layer underneath them: how many generations a pack actually returns, what a retry costs, and how to plan a week of clips without discovering half-way through that the balance is gone.
The cost table, with no rounding up
Every number below is the amount the studio deducts, not an estimate:
| Setting | Credits per clip |
|---|---|
| Text or image to video, 5s, no audio | 300 |
| Text or image to video, 5s, with audio | 600 |
| Text or image to video, 10s, no audio | 600 |
| Text or image to video, 10s, with audio | 1,200 |
| Motion control, standard 720p | 500 |
| Motion control, pro 1080p | 1,000 |
The entry pack is 600 credits for $9.90. So the useful framing is not "how many credits do I get" but "how many finished videos does 600 credits buy me at the settings I actually use".
At 300 credits a clip: two. At 500: one, with 100 left over. At 600: one. At 1,000: none — you need two packs for a single pro motion-control clip.
That last row is why people are surprised. Pro motion control is not "a bit more"; it is more than the entry pack.
The retry multiplier
A generation is not guaranteed to produce a usable clip. Two different costs hide here:
- A refused job (file too large, no detectable face, policy stop) does not deduct credits. It costs time, not money.
- A completed job with a bad result deducts full credits. The model ran; the cost was real.
So the planning unit is not "clips" but "attempts per usable clip". A realistic range:
| Experience level | Attempts per usable clip | Effective cost per usable clip |
|---|---|---|
| First session, vague prompts | 3 | 900 credits |
| After reading the prompt guidance | 2 | 600 credits |
| Proven photo + proven prompt template | 1.2 | 360 credits |
The first row is why someone spends $9.90, gets one usable clip, and concludes the tool is expensive. The third row is the same tool with the inputs fixed. The difference is a photo, a sentence, and one avoided variable per retry.
Budgets should use 2 as a planning default until your own data says otherwise. That means one 600-credit pack is roughly one publishable clip for a new user, not two.
Three weekly plans, priced
Plan A: one clip, one occasion
A birthday message for a family group chat. No public post, no trend, no sound page.
- 1 idea × 2 variants × 300 credits = 600 credits (one pack, $9.90)
- Expected: one good clip, one spare.
This is the honest minimum purchase. If the first attempt is good, the second variant is a free upgrade; if it is not, you are not re-buying.
Plan B: three posts a week
Three short clips for a small account, posted over a week.
- 3 ideas × 2 variants × 300 credits = 1,800 credits (three packs, $29.70)
- Add music from the platform library: 0 extra
- Add one motion-control clip at standard 720p instead of a text clip: +200 credits
Three packs for three posts, not one pack for three posts. The difference between those two expectations is the single most common cause of a frustrated first month.
Plan C: ten clips for a launch week
A small product or a themed week — say ten clips across seven days.
- 10 ideas × 2 variants × 300 credits = 6,000 credits (ten packs, $99)
- Or one 50,000-credit pack, which costs the same as ten Try packs and returns over eight times the credits.
That last comparison is worth reading twice. Ten Try packs cost $99 and return 6,000 credits; the $99 Max pack returns 50,000 — more than eight times the credits for the same money. If you are doing a ten-clip week, you are already in that bracket, and the smaller packs are the expensive way to get there.
Run your own numbers before you decide:
The calculator reads the same table the studio deducts from, so the plan it prints is the plan you will be charged.
Where credits leak
Five leaks, in descending order of cost:
- Ten-second generations to fix a five-second problem. Double cost, more drift, and the fix was a trim.
- Generated audio plus a platform song. You pay double and bury the audio.
- Retrying an identical prompt. Same input, same defect class, same charge.
- Pro motion control for a photo problem. The mode transfers motion faithfully onto a photo that still has a small face in it.
- Regenerating for a crop. Cropping is free in any editor; the frame is a prompt clause, not a re-render.
None of these are model failures. They are planning failures, and each one is avoidable by a sentence or a habit.
Sequencing a week without regenerating
The efficient pattern is not "generate, publish, generate, publish". It is:
- Generate silent variants in one sitting. Batches of two per idea, 300 each.
- Trim all of them the same day, to the accent frame. Trimming is free and finds the clips that do not work before you plan around them.
- Attach the sound at upload, from the platform library, per platform.
- Keep every silent export, including the ones you did not publish.
Step 4 is the compounding part. A library of silent, trimmed clips can be re-cut with a new sound when a trend returns, at zero credits. Regenerating a clip because the old one had baked-in audio is 300 credits every time.
When the arithmetic says "do not buy"
There are cases where the honest answer is that a paid clip is the wrong purchase:
- You only want to see what it looks like. That is a research question, not a generation. Read the mode comparison and the failure guide first.
- You do not have a photo you have the right to use and you do not want a generated subject. Then the format is not available to you, and no pack fixes it.
- You are planning twenty posts from one idea. One clip will not carry twenty posts; that is a content problem, not a credit problem.
Buying 600 credits to answer a question that a page could answer is the most common waste in this category, and it is the one this whole guide set exists to prevent.
The summary arithmetic
- Cheapest usable clip: 300 credits (silent, 5s, text or image mode).
- Realistic attempts per publishable clip for a new user: 2.
- So a 600-credit pack is one publishable clip, or two if your inputs are already right.
- A ten-clip week is roughly 6,000 credits: two Pro packs (3,000 credits each, $29) cover it for $58, where ten Try packs would cost $99.
If that arithmetic changes your plan, it has done its job. If you are still deciding which mode to spend on, the mode comparison is the next thing to read — it is free, and it is the decision that sets every number above.
Frequently asked questions
How many clips do I get for $9.90?
The Try pack is 600 credits. At the cheapest setting — a five-second text-to-video or image-to-video clip without audio at 300 credits — that is two clips. With generated audio, or at ten seconds, one clip costs 600 and the pack returns one. At motion-control pro (1,000 credits) a pack does not cover a whole clip.
Do I need to budget for failures?
Yes, in the sense that a retry is a full-price job. A failed job that the provider rejects is not charged, but a completed job with a bad result is. If you expect to retry once per usable clip, budget two generations per published clip.
What is the cheapest way to make a batch of clips?
Generate silent five-second clips at 300 credits each, generate two variants of each idea rather than one, and add music from the platform library in the editor. Silent clips are the only setting where a 600-credit pack reliably returns two finished videos, and they can be re-cut later for free.
Is a bigger pack better value?
Per credit, yes — the packs scale and larger ones cost less per credit. The question is whether you will use the credits before you lose interest or the trend fades. Credits have no scheduled expiry on the one-time packs, so the risk is unused balance, not forfeiture.
Can I plan a week of content without generating anything?
Yes, and you should. Write the ideas, count the clips at the settings you will use, price the retries, and only then buy. The calculator above does the arithmetic against the same deduction table the studio uses.
![50+ Best AI Baby Dance Video Prompts That Actually Work [2026 Guide]](/imgs/blog/baby-dance-ai-prompts.webp)
![AI Baby Dance Video Generator: How to Make Your Baby Dance with AI [2026]](/imgs/blog/how-to-make-baby-dance-ai-video.webp)







