Wan 3.0 Document to Video

Drop in a deck, a report or a spreadsheet. Wan 3.0 reads all of it, decides what the story is, and generates new footage with sound in one pass of up to thirty seconds. Not your slides with a voice over them — footage that did not exist before. Your file is checked in the browser first, so one that cannot run never costs you anything.

Send a document

Your document

docx · doc · xlsx · xls · pptx · ppt · pdf · txt · key · pages · numbers · md

Holding a link rather than a file? Wan 3.0 reads public pages too — that runs on the URL to video page →

What should Wan 3.0 take from it?

0 / 20,000

Document to video runs on credits. The free clip is a prompt or a photo in, video out — reading a file, a link or footage you already have is the paid side of the same Wan 3.0 model. See what a clip costs.

480P · 16:9 · 20s · ×1
The document costs nothing. You are billed for output seconds only.
If it fails, the credits come back. Automatically, no ticket.

1 free clip on the real Wan 3.0 · the document itself is billed at zero · failed runs cost nothing

Every idea on the site, from all six tools · press Use this prompt to load one into the panel on the left

Five dancers in black dresses silhouetted against a hard white backlight in ground fog

12s · 16:9 · 720P

Dancers in backlit fog

What comes back

What Wan 3.0 document to video gives you back

Wan 3.0 document to video takes one file — a deck, a report, a spreadsheet — reads all of it, works out what the material is about, and directs a new video from that understanding. Picture and sound together, up to thirty seconds, at 480P, 720P or 1080P. It reads your document. It does not put it on screen.

What Wan 3.0 is not: a slideshow exporter, or a narrator reading your bullets over your layout. If the deck's design is the thing you ship, a section below says so plainly.

One continuous take: 2–30 s
100 MB · one file per run: 50 pages
Deck · doc · sheet · PDF: 12 formats
What the document costs: $0

Not a slideshow exporter
Not a narrator over your layout
Not one scene per slide

Free · no account · nothing uploaded

Will your file run, and what fits in thirty seconds

Two questions decide whether a Wan 3.0 document job is worth starting, and both are answerable before you spend anything. These checks run in your browser; nothing is uploaded until you press generate.

Your file, checked before Wan 3.0 sees it

docx · doc · xlsx · xls · pptx · ppt · pdf · txt · key · pages · numbers · md

Formatone of the twelve
Size100 MB, whatever the format
Pages50 — a count, not a size
Nothing else attachedno first or last frame

Spreadsheets, txt and md are size-checked only: Alibaba validates the fifty-page ceiling for pdf, docx, doc, pptx, ppt, key and pages, and holds the rest to 100 MB alone. A nine-hundred-row sheet is not nine hundred pages, and Wan 3.0 will not refuse it for length.

Seconds per page

Here is the arithmetic that decides whether you end up disappointed. Fifty pages into thirty seconds is 0.6 seconds a page, and nobody reads at that speed — so Wan 3.0 does the only sane thing left and makes a trailer.

Pages in5s10s20s30s

12 pages: 0.41; 0.83; 1.66; 2.50
34 pages: 0.14; 0.29; 0.58; 0.88
50 pagesthe ceiling: 0.10; 0.20; 0.40; 0.60

Under roughly a second a page you are commissioning a teaser, so cut the deck, not the clip. Twelve pages of the same material beats the full fifty, because the selecting was done by you rather than delegated to Wan 3.0. Cannot cut it? Write the conclusion as a prompt instead.

Three consequences

What Wan 3.0 does with your pages

Three things follow from how Wan 3.0 works — it reads the document, then directs a film. None is a defect, and all three are easier to plan for than to discover afterwards.

Structure is read, not kept
Your section order is a signal, not a storyboard. A document and a prompt travel in the same Wan 3.0 request, and the prompt is where you steer.
Put the conclusion in the first third
Density is flattened
Thirty seconds cannot carry fifty pages, so the model selects. What it picks is not always what you would have, and that is the usual reason a first attempt disappoints.
Name the exclusions in the brief
Numbers need a second look
Data-heavy sources are the least predictable input. Values come through; the furniture around them does not.
Read every figure against the source

None of the three is a reason to avoid a document. They are the reason to say what you want out of it, which is what the next section builds.

The exclusions do half the work

Tell Wan 3.0 what to take and what to ignore

A Wan 3.0 brief for a document is not shaped like a text-to-video prompt. There, the description does all the work. Here the exclusions do half of it — your file always holds more than thirty seconds can carry, and if you do not say what to drop, the model drops something for you.

Take

The facts that have to survive. Names, figures, dates, the one claim.

Ignore

The parts of the file that are not this video's job.

Treatment

Sound

Length

Your brief

A 20-second product launch film built from the attached document. Take the product name, the three headline capabilities and the launch date. Ignore pricing, the roadmap and anything about the team. Style: clean studio product photography, cool neutral palette, one slow push-in per beat. Confident voiceover, light ambient bed, no music sting.

Want the full prompt structure in Alibaba's own order? The builder writes it →

Every number, with its source

Which files Wan 3.0 accepts, and the two ceilings

Document formats: docx, doc, xlsx, xls, pptx, ppt, pdf, txt, key, pages, numbers, mdTwelve, and nothing else; API ref
File size: 100 MBWhatever the format; API ref
Page count: 50Validated for pdf, docx, doc, pptx, ppt, key, pages; API ref
Per request: One file, or one link. Never both, never several; API ref
A link instead: One public page can take the file's placeIt has its own page · /url-to-video; API ref
Output: 2–30 seconds at 480P, 720P or 1080P4K: not supported; API ref
Audio: Generated in the same pass as the picture; API ref

The one that catches people out: the page ceiling is a page count, not a proxy for size. A slim eighty-page PDF is refused by Wan 3.0 while a heavy forty-page deck goes through. And the ceiling is only validated for seven of the twelve formats — a spreadsheet is held to 100 MB alone. Source: Wan 3.0 video generation API reference, Alibaba Cloud Model Studio, re-checked 27 August 2026.

Before this faces a client

Your numbers survive. Your chart furniture does not.

This decides whether a Wan 3.0 clip can face a client, and it splits cleanly in two. The figures come through. What the model draws around them is where it falls apart.

Holds up

Cell values and percentages
Product names and short slogans
Dates and single headline figures
One number, large, on its own

Comes apart

Legends — two series can fuse into one label
Axis ticks — equal gaps, unequal values
Thousands separators and decimal points
Anything carrying more than two series

The fix is not a better Wan 3.0 prompt. It is a smaller ask.

01
Write each figure as a quoted literal — "$1,580", not "the January number". You are handing Wan 3.0 a string to copy, not a fact to look up.
02
Ask for oversized standalone numerals and one coloured badge instead of a chart.
03
Delete legends, axes and tick labels from the brief. Nothing left to draw is nothing left to garble.

That split comes from an independent frame-by-frame audit of Alibaba's official Wan 3.0 spreadsheet demo, August 2026. We have not reproduced it ourselves — our own document runs will sit here with their prompts underneath.

Four kinds of file

Which source suits which Wan 3.0 video

Wan 3.0 takes all four, but not equally well. Alibaba ships these as its document use cases; the last column is the part it leaves out.

The fileGood forCheck afterwards

A slide deck
Launch films, brand pieces, conference teasers
Check afterwards
product shape and colour across shots — a physical object can drift between beats
A report or PDF
Narrated briefings, summary clips
Check afterwards
quotes, attribution, and whether the argument survived compression
A spreadsheet
Trend pieces, one-figure stories
Check afterwards
every number, every label, every unit
A course or manual
Lessons, explainers, onboarding pieces
Check afterwards
whether a character or prop drifts between beats

One document, one idea, one video. A deck covering a product launch and a hiring pitch gives Wan 3.0 nothing to commit to. Paste the speaker notes into the brief — they carry meaning the slides do not — and send the summary sheet, never the raw data.

The rule that refuses a run

One file or one link — never both, never several

Wan 3.0 sorts media inputs into two families and one request may draw from only one. Documents and links sit in the reference family with reference images, video and audio. First and last frames are the other. Send one of each and the request is rejected outright, so this page greys out the frame inputs the moment a document loads.

Reference family

file — one document per run
link — one public page, instead of the file
reference_image
reference_video
reference_audio

Frame family

first_frame
last_frame

Pick one column. Sending both is refused before anything generates.

Inside the left column, a file and a link are also one or the other, so you cannot hand Wan 3.0 the deck and the product page in the same run. One document per run sets the real cost here: change one figure and the whole clip regenerates, because there is no way to re-render shot three on its own. Which is the argument for putting the first pass through at 480P.

Read this before you pay

When Wan 3.0 document to video is the wrong tool

Worth being direct, because a good share of the people searching for this want something else. If you need your slides on screen, in your layout, with a narrator reading them — compliance modules, onboarding, multilingual training — that is a different product category, and the tools built for it keep your layout and add a voice. Wan 3.0 does the opposite: it discards the layout and generates new footage.

The rule of thumb is short. If the design of the deck is what you are shipping, do not use a generative model. If the deck is source material for something that will look nothing like it, Wan 3.0 has almost no competition.

Not your layout, kept
Not a voice over your bullets
Not a slide-by-slide export

Pricing · Native audio · No watermark

What a Wan 3.0 document to video clip costs

The document is free. Wan 3.0 bills output seconds at the resolution you choose, and nothing is added for the file, its page count, its size, or a link in place of it. Twenty seconds from a fifty-page deck costs what twenty seconds from one line of text costs. And a failed run is refunded automatically.

How this is billed

Billed by: output second
The document: free
Its page count and size: free
Failed generation: refunded
Your first clip: free · 480P

Put the first pass through at 480P. Wan 3.0 document jobs miss more often than prompt jobs, because the model is making editorial choices you have not watched it make before — and one file is one run, so a re-cut is a whole regeneration. The full credit table

Same monthly credits either way

Free first clipNo card
One generation, once, on the real wan3.0-video. A sample rather than an allowance — every free second is cash spent before any arrives.
$0once
No card at any point
Generate free clip
An email, no card
1 clip, once
480P · up to 3s · with sound
An email, no card
Wan 3.0 has no free tier of its own — Alibaba meters it from the first second
- The real wan3.0-video, not 2.7
- Sound generated with the picture
- Model ID and task ID on the receipt
- A failed run never costs you
- 720P and 1080PPaid
- Clips longer than 3 secondsPaid
- Reference, document and editing modesPaid
- More than one clip, everPaid
Anyone advertising unmetered free access is either serving a different model or paying a bill that will not last.
At a glance
Clips a month
1, once
Takes per run
1
Longest clip
3s
Top resolution
480P
StarterSave 50%
A launch post a week. Enough to find out whether the Wan 3.0 AI video generator suits the way you work, without a decision about volume.
$9.90/mo$19.80
$118.80 billed yearly
Cancel anytime
480 credits / month · ≈ 12 clips
at 5s 480P · 6 at 720P · 3 at 1080P
Plan credits expire at the end of the month
- **480 credits a month** — about 12 finished clips
- 2 takes of one prompt per run
- Every resolution — 480P, 720P and 1080P
- Every length — 2 to 30 seconds in one pass
- Every input — text, image, reference, document, web page
- Sound written in the same pass, no watermark
- Commercial use included — publish it, sell it, bill for it
- Wan 3.0 Prime, the fast tier, whenever you want it
- A failed run never costs credits
- 7-day refund while the credits are untouched
- Cancel in two clicks — the price you joined at is locked
Everything under the first two lines is on this plan at $9.90. The bigger plans buy seconds and takes — never a better model.
At a glance
Clips a month
≈ 12 at 5s 480P
Takes per run
2
Credits per dollar
Baseline
Refund window
7 days
Most people land here
Pro
One finished clip every working day, with three alternates each. The draft-at-480P, finish-at-1080P loop, sized for a working week.
$29.90/mo$59.80
$358.80 billed yearly
Cancel anytime
1,760 credits / month · ≈ 44 clips
at 5s 480P · 22 at 720P · 11 at 1080P
Plan credits expire at the end of the month
- **1,760 credits a month** — 3.7× Starter4×
- ≈ 44 finished clips a month, or 11 at full 1080P
- **4 takes of one prompt per run** — pick one instead of re-rolling2× takes
- **+32% credits per dollar** than StarterBetter rate
- Room to draft at 480P and finish at 1080P in the same week
- Every resolution — 480P, 720P and 1080P
- Every length — 2 to 30 seconds in one pass
- Every input — text, image, reference, document, web page
- Sound written in the same pass, no watermark
- Commercial use included — publish it, sell it, bill for it
- Wan 3.0 Prime, the fast tier, whenever you want it
- A failed run never costs credits
- 7-day refund while the credits are untouched
- Cancel in two clicks — the price you joined at is locked
The jump from Starter is quantity, takes and rate. The model, the resolutions and the thirty seconds were already yours.
At a glance
Clips a month
≈ 44 at 5s 480P
Takes per run
4
Credits per dollar
+32% vs Starter
Refund window
7 days
StudioBest rate
Client volume, and the plan where you stop rationing 1080P. Forty full-resolution finished clips a month, at the lowest credit price on this page.
$79.90/mo$159.80
$958.80 billed yearly
Cancel anytime
4,960 credits / month · ≈ 124 clips
at 5s 480P · 62 at 720P · 31 at 1080P
Plan credits expire at the end of the month
- **4,960 credits a month** — 10× Starter, 2.8× Pro13×
- ≈ 124 finished clips a month, or **31 at full 1080P**
- **+65% credits per dollar** — the best rate on this pageBest rate
- Enough 1080P that you stop drafting at 480P first
- 4 takes of one prompt per run
- Sized for a client roster rather than one channel
- Top up any month without changing plan
- Every resolution — 480P, 720P and 1080P
- Every length — 2 to 30 seconds in one pass
- Every input — text, image, reference, document, web page
- Sound written in the same pass, no watermark
- Commercial use included — publish it, sell it, bill for it
- Wan 3.0 Prime, the fast tier, whenever you want it
- A failed run never costs credits
- 7-day refund while the credits are untouched
- Cancel in two clicks — the price you joined at is locked
Studio is volume and rate, nothing else. If you are not spending a month’s credits, Pro is the honest answer — we would rather write that here than take the difference.
At a glance
Clips a month
≈ 124 at 5s 480P
At full 1080P
≈ 40 a month
Credits per dollar
+65% vs Starter
Refund window
7 days

Before you upload the quarterly

Where your document goes after you upload it

Launch decks, quarterly numbers and internal proposals are exactly the files people want as video, and exactly the files with a policy attached. So, plainly: a document sent here leaves your building. It goes to Alibaba Cloud, which runs Wan 3.0, for as long as the job takes.

Nothing is uploaded by the checkFormat, size and page count are read in your browser. The file only leaves your machine when you press generate.
It goes to Alibaba CloudThat is where Wan 3.0 runs. For as long as the job takes, and no longer.
We train nothing on itNo model of ours reads it, and it appears in no gallery or example anywhere on this site.
Our copy is deletedOnce your clip is delivered, the stored file goes with it.

If the file is under an agreement that keeps it in-house, the honest answer is that no hosted model can help you — Wan 3.0 included.

The binding version is on the privacy policy →

Questions people actually ask

Wan 3.0 document to video questions

Ten answers on Wan 3.0 document to video · all visible · nothing collapsed

Can I actually upload a document here?

Yes. Drop a file at the top of this page, say what to take from it, and generate. Every account gets one free clip on the real Wan 3.0, and the format, size and page checks run before anything is sent.

What file types can Wan 3.0 turn into video?

docx, doc, xlsx, xls, pptx, ppt, pdf, txt, key, pages, numbers and md, up to 100 MB and 50 pages. One file per run — a public page can take a file's place, but that has its own page.

Does Wan 3.0 turn each slide into a scene?

No, and this is the most common misunderstanding. It reads the whole document and directs a video from what it learned. Fifty pages into thirty seconds is 0.6 seconds a page, so you get a trailer, not a page-turn.

Does the 50-page cap apply to my spreadsheet?

Not as a page count. Alibaba validates that ceiling for pdf, docx, doc, pptx, ppt, key and pages; spreadsheets, txt and md are held to the 100 MB limit alone.

Can I use a document and a first-frame image together?

No. Files and links belong to the reference family, which is mutually exclusive with frame inputs. Wan 3.0 rejects a request carrying both before anything generates.

Does a bigger document cost more in Wan 3.0?

No. Output seconds are the only billed thing, so a fifty-page deck and a one-line prompt cost the same for the same length of clip.

Are the numbers from my spreadsheet accurate in a Wan 3.0 clip?

Cell values generally survive. The furniture breaks — legends merge, axis ticks stop being linear, separators go missing. Quote every figure as a literal and design charts out of the brief.

Why did Wan 3.0 ignore the most important part of my deck?

Usually because it sat late in the document and the brief did not name it. Move the conclusion into the first third and put what you care about in the Take column.

I have a link, not a file. What then?

Wan 3.0 reads a public page the same way it reads a document, but a link brings its own set of failure modes, so it gets its own page and its own preflight. Start there instead.

Can Wan 3.0 make a training video from our manual?

It can make a thirty-second piece *about* the manual. If you need the procedure taught step by step in your layout, that is narrated-slide tooling rather than a generative model.

Send a deck through Wan 3.0

Upload one file, say what to take from it, and watch what Wan 3.0 does with the rest. One free clip at 480P, up to three seconds — an email, no card. The document is billed at zero, and a failed run is refunded without you asking.

Written and maintained by the wan-3.run editorial teamPublished 27 August 2026Last updated 27 August 2026