Text to video
Fastest startWrite the idea or paste a script and the model decides the look, framing and motion. Best for exploring an idea quickly.
Full videos up to 20 minutes, with voices and music
The AI video maker that turns your idea into a finished video
Most AI tools give you a few seconds of footage and leave the rest to you. Describe your video here, or start from a photo, and get the whole thing back: planned shots, consistent characters, voices, music, and an MP4 ready to post.
Your prompt
A one-minute video about a small café opening for the day. Warm, cinematic, with a calm narrator.
Planned shots
- 0:00Sunrise over the harbour, slow push in
- 0:12Maya opens the café and flips the sign
- 0:31First customer arrives, Maya smiles
- 0:48Close-up of the coffee, warm light
VoicesMusicSame charactersMP4 download
- Longest single video
- 20 minLongest single video
- AI models to choose from
- 3AI models to choose from
- Aspect ratios
- 6Aspect ratios
- Watermarks
- 0Watermarks
How it works
How to make a video with AI in four steps
Only the first step is required. The others are there when you want more control.
- 1
Say what the video is about
Type the idea in plain language: who is in it, what happens, where, and how it should feel. One sentence is enough to start, and a full script works too.
- 2
Add images if you have them
Upload a photo as the opening shot, a second one as the closing shot, or reference pictures of a person, product or character you want to keep the same throughout.
- 3
Pick the length and format
Choose anything from a few seconds to 5 minutes (up to 20 on Pro), then 16:9 for YouTube, 9:16 for Shorts, TikTok and Reels, or 1:1 for feeds. Leave sound on to get voices, effects and music.
- 4
Generate, review, download
Your video renders in the background, and you see the credit cost before it starts. Download the MP4 when it is ready, or change one detail and run it again.
Making a video used to mean a script, a shoot, an edit and a sound mix, usually over days and usually with several people. An AI video maker folds those stages together. The planning, the footage and the audio all come from the same description, so the slow part of the job moves from production to deciding what you want to say.
It also changes how you work. When one version costs minutes rather than a day of filming, you can try a few directions and keep the best, rather than committing to a single idea before you have seen it.
Start here
Prompts you can use right now
Each card is a complete prompt. Open it in the studio, change the details, and make it yours.
Lifestyle ad
15s9:16
Coffee brand ad
Lifestyle adClose-up of dark espresso pouring into a white cup in a sunlit café, steam curling upward, then a barista slides it across the counter with a smile. Slow push in, warm light, soft jazz.
Cinematic
30s16:9
Short film opening
CinematicA lighthouse keeper climbs a spiral staircase at dusk, lights the lamp and watches a storm roll in over black water. Slow crane up, cinematic film look, distant thunder.
Product studio
10s1:1
Product reveal
Product studioA white running shoe rotates on a dark glossy surface as a line of light sweeps across it, then splashes of water freeze in mid-air around it. Studio lighting, slow orbit, deep bass hit.
3D explainer
60s16:9
Science lesson
3D explainerA cutaway of a volcano shows magma rising through a chamber, pressure building, and ash bursting from the top. Clean 3D explainer, calm narrator describing each stage.
Anime
20s16:9
Anime short
AnimeA girl on a bicycle races a summer storm along a coastal road, her scarf whipping behind her as the first rain falls. Cel-shaded anime, dynamic tracking camera, soaring strings.
Documentary
60s16:9
Faceless history video
DocumentaryA sweeping shot of a medieval port city at dawn, then ships unloading spices as merchants argue on the dock. A calm narrator introduces the rise of the city. Cinematic, natural light.
Four ways in
Four ways to start an AI video
Text is the quickest. The other three are for when you need the result to match something specific.
Image to video
Keeps your pictureYour image is the first frame, so faces, products and logos stay as they are. Write about what happens next rather than what it looks like.
First and last frame
Controls the endingGive an opening and a closing image and the model works out the journey between them. Good for transitions, reveals and loops.
Reference to video
Same subject, new sceneAttach pictures of a person, product or place and put them in a scene those pictures never showed. This is how a series or campaign stays consistent.
Two more modes work on video you already have. Extend continues a clip past its last frame, and Edit changes something within a clip while leaving the rest alone. See the AI video editor guide for when each one helps.
Which one should you use?
If you have only an idea, start with text and keep the first attempts short and at 480p. You are testing the concept, so spend as little as you can.
If something has to look a certain way, such as your product, your face or your artwork, use image to video so that it starts from the real thing.
If you are making more than one video with the same subject, use reference images. It takes a minute to prepare and saves you from describing the same character in every prompt.
How to write an AI video prompt
The biggest improvement comes from describing a shot instead of a topic. Compare these two:
Weak: “A video about coffee.”
Strong: “Close-up of a barista pouring steamed milk into a dark espresso in a small, sunlit café. Steam rises as the pattern forms. Slow push in, warm morning light, gentle acoustic music.”
The second one covers everything the model needs. Anything you leave out gets filled with the most average answer, which is why vague prompts give generic results.
The five things to include
- Subject: who or what we are looking at, with one or two specific details.
- Action: what happens, in order. Motion comes from verbs.
- Setting: the place, the time of day and the light.
- Camera: for example a static wide shot, a slow push in or a handheld follow.
- Look and sound: the visual style, the mood and what we hear.
For spoken words, put the line in double quotation marks and it will be treated as dialogue. For a long video you can write the idea once and let the tool plan the shots, or paste your own script. There are ready-to-use examples in our AI video prompt library.
Look and feel
One tool, any visual style
An AI video maker has no house style. The look comes from the words you use, so name it on purpose.
| Style | Words to use | Good for |
|---|---|---|
| Cinematic | cinematic film still, anamorphic lens, natural light, shallow depth of field | Trailers, brand films, drama |
| Documentary | handheld documentary footage, available light, natural colour | Stories, nature, real-world topics |
| Product studio | studio product shot, soft box lighting, seamless backdrop, slow orbit | Ads, ecommerce, launches |
| 3D animation | stylised 3D animation, soft global illumination, rounded forms | Mascots, kids' content, tech brands |
| Anime | cel-shaded anime, expressive eyes, dynamic camera | Story channels, gaming, music |
| 2D cartoon | 2D cartoon, bold outlines, flat colour, snappy timing | Comedy, explainers, social |
Pick one style and stay with it. Mixing words from two different looks tends to produce something that is neither. Name the colour palette and the lighting separately, because those two details shape the mood more than almost anything else.
See every option, each with a starter prompt, on the video styles page.
Keeping characters and products consistent
Consistency is where most AI video falls apart. A description such as “a friendly woman in a blue jacket” fits thousands of different people, so each new shot may land on a different one.
Three things help, in order of strength. Specific details in the prompt narrow the range. Starting from an image fixes the look for that shot. Reference images do the most, because the model has an actual picture to hold on to in scenes it has never seen.
When you make a long video or a series here, the characters are planned once at the start and carried through every shot, rather than reinvented each time. The consistent characters guide explains how to prepare good reference images.
Voices, music and sound effects
Many AI video tools produce silent footage, which means a separate job in an editor afterwards. Here the sound is part of the generation. Dialogue is lip-synced, effects land on the action, and the music follows the mood of the scene.
To control it, put spoken lines in double quotation marks, and describe what we should hear as well as what we should see: “rain on a tin roof, distant traffic, soft piano”. Keep each line short, about two or three seconds. If you plan to add your own soundtrack, you can turn audio off.
Duration
How long should your video be?
Choose the length that suits the job. A short video that ends well beats a long one that drags.
| Type of video | Good length | Why |
|---|---|---|
| Ad hook test | 6–15s | You only need to learn whether people stop |
| TikTok, Reel or Short | 15–45s | One idea, finished before attention drops |
| Product video | 20–40s | Problem, product, result, and out |
| Explainer | 1–2 min | Enough room for one idea done properly |
| Story or lesson | 3–5 min | Space for a beginning, middle and end |
| YouTube episode | 8–20 min | Long enough for real watch time |
This is where we differ from most tools. A single AI clip is normally limited to between 8 and 30 seconds, and joining clips up by hand is where stories fall apart. Here you pick 1, 2, 3 or 5 minutes, or 10 or 20 on Pro, and receive one finished video planned as a whole. The long AI video generator guide goes into detail.
Formats
Choose the right shape for where it will play
Pick this before you generate. The shot is composed for the frame, so cropping later throws that away.
| Where it plays | Aspect ratio | Note |
|---|---|---|
| TikTok, Reels, Shorts | 9:16 | Full-screen vertical |
| YouTube | 16:9 | The standard for long video |
| Instagram and Facebook feed | 1:1 or 3:4 | Takes up more of the screen |
| Website banner | 21:9 or 16:9 | Wide and generated to fit |
| Presentation | 16:9 | Matches a standard slide |
| In-store screen | 9:16 or 16:9 | Check the screen's orientation first |
If you start from your own image or video, the result takes that file's shape, so prepare the image in the format you want. For tall video, see the short video maker guide.
Who it is for
Who uses an AI video maker
The tool is the same, but the job is different. Each link below goes to a guide for that job.
YouTubers
Publish regularly without filming: long videos, Shorts and faceless formats.
AI YouTube video maker →Marketers
Ad variations, launch videos and social clips, ready the same day you think of them.
AI ad video maker →Online sellers
Show a product in use, in any setting, starting from the photos you already have.
AI product video maker →Teachers and trainers
Lessons and onboarding that show the idea rather than read it out.
AI training video maker →Short-form creators
Daily vertical clips, with hooks built for the first two seconds.
AI short video maker →Storytellers
Short films and series with characters who stay the same from episode to episode.
AI series generator →What people make
What you can make with AI video
The best uses share one thing: filming them would be slow, expensive or impossible.
Ads and promos
Try several hooks for a product before spending on a shoot, then scale the one people click.
YouTube channels
Full videos and Shorts for channels that run without a camera, a studio or a presenter.
Short-form social
Vertical clips for TikTok, Reels and Shorts, composed for a tall frame rather than cropped into it.
Lessons and training
Turn a process or a lesson into a scene, including things a camera could never film.
Stories and series
Short films and multi-episode stories where the same characters come back every time.
Bringing images to life
Make an old family photo, a product shot or a piece of artwork move.
Where AI video is the right choice
When you need volume. A channel, an ad account or a social feed needs regular new material, and that is hard to keep up with a camera.
When the scene cannot be filmed. The inside of a cell, a city in 1890, a product that does not exist yet. These cost nothing extra to describe.
When you want to test before you commit. Try an idea as a rough video, see the reaction, then decide whether it deserves a real production.
And where it is not
Some things should stay real. When the message depends on trust, such as a founder's apology, a customer testimonial or a review, a real person on camera is the right choice, and a generated one would mislead people. The same goes for footage presented as a record of real events. Be open about AI-generated content, and follow the disclosure rules of the platform you post on.
Comparison
AI video maker vs filming it yourself
Both have their place. This is what each one is actually good at.
| AI video maker | Traditional shoot | |
|---|---|---|
| Time to first version | Minutes to about an hour | Days to weeks |
| What you need | A clear description | Camera, crew, location, editor |
| Changing your mind | Edit the prompt and run it again | Often means a reshoot |
| Impossible scenes | Same cost as any other scene | Expensive visual effects |
| Real people and places | Should not be faked | The right tool |
| Exact on-screen text | Add it afterwards | Handled in the edit |
| Frame-perfect control | Not yet | Complete |
| Best at | Trying many ideas quickly | One definitive piece |
The real change is not only the price of one video. It is that testing several versions of an idea no longer needs a budget, so you can let the results decide which one deserves more.
Troubleshooting
Fixing the most common AI video problems
Change one thing at a time and run it again, so you can see which change made the difference.
| Problem | Likely cause | What to do |
|---|---|---|
| The character changes between shots | The description fits too many people | Add reference images or start from a picture |
| The camera drifts for no reason | No camera direction in the prompt | Add “static camera” or “locked-off shot” |
| The video feels empty | Too much length for one idea | Shorten it, or add another event |
| Lips do not match the speech | The line was described, not quoted | Put the words in double quotation marks |
| It looks generic | No style, light or palette named | Name the look and the lighting |
| Words on signs look wrong | Generated text is unreliable | Ask for blank signs and add text afterwards |
What an AI video costs
Videos are priced in credits, and you see the cost before you start. Three things set it: the length, the model and the resolution. A 5-minute video costs ten times a 30-second one, and 480p costs half of 720p. If a render fails, the credits are refunded automatically.
The figure to plan around is the cost of a video you actually keep. Expect to run a few attempts. Draft short at 480p, and only render the full-length version once the idea works. See pricing for plans and one-off packs.
What AI video does well, and what it does not
It is strong at people and characters in motion, natural scenes, products, atmosphere and storytelling, and it is quick enough for you to try several ideas.
It is weaker at written words on screen, at hands doing intricate work, at matching a pre-recorded voice track frame by frame, and at keeping tiny details identical over many shots. Plan around these: add text afterwards, avoid close-ups of fingers doing delicate jobs, and use reference images for anything that must stay the same.
Some things you should not do at all, whatever the tool allows, such as depicting real people without consent or passing off generated footage as real events. Our terms set out what is not allowed.
Before you publish
An AI video checklist
Ten quick checks that catch nearly every embarrassing mistake.
- Watched it all the way through with the sound on.
- Watched it again muted, and the point still lands.
- Something happens in the very first second.
- Hands, faces and fine details look right in every shot.
- The same character looks the same at the end as at the start.
- No garbled writing is visible on signs or screens.
- Titles and captions added in an editor where needed.
- Made in the right shape, not cropped afterwards.
- Nothing shows a real person or brand you lack the right to use.
- Downloaded the MP4 and checked it on a phone.
Reference
AI video glossary
The terms used on this site, in plain English.
- Prompt
- The written instruction for a video. Good prompts name the subject, the action, the setting, the camera and the look.
- Text to video
- Generating video from words alone, with no image supplied.
- Image to video
- Generating video that starts from a picture you provide, which becomes the first frame.
- Reference image
- A picture attached to a generation so the model keeps that person, product or style in a new scene.
- First and last frame
- Giving the opening and closing images so the model creates the movement between them.
- Aspect ratio
- The shape of the frame. 16:9 is wide and 9:16 is tall. Pick it before you generate.
- Resolution
- How many pixels the video has. Draft at 480p because it costs half, then render the final version at 720p or above.
- Video extend
- Continuing an existing clip past its last frame.
- Video edit
- Changing something inside a clip you have by describing the change, while keeping the rest.
- Credit
- The unit videos are priced in. Cost depends on length, model and resolution, and failed renders are refunded.
Go deeper
AI video maker guides
Every feature and use case, explained on its own page.
Create
AI video generator
From one sentence to a finished video.
Read →Text to video AI
Write a scene and get footage with sound.
Read →Image to video AI
Bring any still image to life.
Read →Photo to video AI
Animate a photo you already have.
Read →AI video editor
Extend or change a clip by describing it.
Read →Long AI video generator
One finished video of up to 20 minutes.
Read →AI series generator
Episodes that continue one story.
Read →Free AI video maker
What you get at no cost, and the limits.
Read →Styles
All video styles
Every look, with the words that make it.
Read →Cinematic AI video
Film-like light, lenses and movement.
Read →Anime video generator
Cel shading, speed lines and drama.
Read →3D animation maker
Soft rendered characters and worlds.
Read →AI cartoon generator
2D characters that move and speak.
Read →Consistent AI characters
Keep one character across every shot.
Read →Creators
AI YouTube video maker
Long videos and Shorts for your channel.
Read →AI short video maker
Reels, Shorts and vertical clips.
Read →AI TikTok video maker
Hooks built for the first two seconds.
Read →Faceless video maker
Run a channel without going on camera.
Read →AI music video generator
Visuals with a beat and a mood.
Read →Business
AI ad video maker
Test ten ad ideas for the cost of one shoot.
Read →AI commercial generator
TV and online spots without a crew.
Read →AI product video maker
Show your product in any setting.
Read →AI explainer video maker
Make a hard idea simple in a minute.
Read →AI training video maker
Lessons and onboarding, without filming.
Read →Compare
Best AI video generators
The main tools compared, honestly.
Read →Best free AI video makers
What is really free, and the catches.
Read →Canva AI video alternatives
When Canva is not enough.
Read →Runway alternatives
Other tools for creative video.
Read →CapCut AI alternatives
Generate footage, not just edit it.
Read →Synthesia alternatives
Beyond the talking avatar.
Read →HeyGen alternatives
Options for presenter and story video.
Read →InVideo AI alternatives
Other prompt-to-video tools.
Read →Sora alternatives
Where to go now Sora has closed.
Read →Seedance vs Kling vs Veo
The big video models, side by side.
Read →From the blog
Learn to make better AI videos
Prompt examples, tool comparisons and honest guides to what AI video can and cannot do.
Best AI Video Generators in 2026: An Honest Comparison
The best AI video generators in 2026 compared: Veo, Kling, Runway, Seedance, Synthesia, HeyGen, Canva and more. Which to use for realism, avatars, templates or full-length videos.
9 min readBest Free AI Video Makers in 2026: What Is Really Free
Which AI video makers are really free in 2026? What the free plans include, what the catches are, and how to get the most from them without paying.
7 min readCanva AI Video Generator Alternatives: 7 Options for 2026
Looking beyond Canva's AI video generator? Compare alternatives for longer videos, sound, realistic footage and consistent characters.
7 min readRunway Alternatives in 2026: Best AI Video Generators Compared
Compare the best Runway alternatives for AI video: Kling, Veo, Seedance, Wan and complete video makers. What each does better, with no filler.
7 min readCapCut AI Video Generator Alternatives for 2026
CapCut is a great editor, but how good is its AI video generator? The best alternatives for generating footage, plus how to use them with an editor.
6 min readSynthesia Alternatives in 2026: Avatars and Beyond
Compare Synthesia alternatives for training and business video: HeyGen, Vyond, and AI video makers that show scenes instead of a talking head.
6 min readFAQ
AI video maker FAQ
What people usually ask before making their first video.
What is an AI video maker?
An AI video maker is a tool that creates video from a description instead of from a camera. You write what should happen or upload an image, and it generates the footage, the movement and the sound. You do not need to film, edit or animate anything yourself.
Is The AI Video Maker free?
You can start without a card. A new account can make a free 5-second preview so you can judge the quality. Full videos, from 1 minute up to 20 minutes, are made with a plan or a one-off credit pack.
How long can an AI video be here?
From a few seconds to 5 minutes on any paid option, and up to 20 minutes on Pro. Most AI video tools stop at a clip of 8 to 30 seconds. Here the whole video is planned from start to finish, so the story and the characters carry through.
Can it make a video from text?
Yes. Text to video is the fastest way in. Describe the scene, or paste a whole script, and the tool plans the shots and renders them with sound.
Can it turn a photo into a video?
Yes. Upload an image and it becomes the first frame, so the person, product or place keeps its appearance while the scene starts to move. Then describe what should happen next.
Does the video come with sound?
Yes. Dialogue, sound effects and music are generated along with the picture, so a door slam lands when the door closes. Put spoken lines in double quotation marks and they are read as speech.
Can I keep the same character or person across a video?
Yes. Attach two or three clear reference images and the model carries that look into new scenes. It is the most reliable way to keep a character the same. Long videos and series keep their characters from shot to shot.
Can it make realistic video, or only animation?
Both. You choose the look in your prompt, from cinematic live-action style to 3D animation, anime or 2D cartoons. See the styles page for the words that produce each one.
Do I need editing experience?
No. There is no timeline. The useful skill is describing clearly what you want to see, and the guides on this site teach that in a few minutes.
How long does a video take to render?
It depends on length, model and how busy the queue is. A short clip takes a few minutes and a long video takes longer. It runs in the background, so you can close the page and we email you when it is ready.
What quality do I get?
MP4 files at 480p for cheap drafts, 720p as standard and 1080p on the models and plans that support it.
Which formats are available?
16:9 for YouTube and websites, 9:16 for Shorts, TikTok and Reels, 1:1 for feeds, plus 4:3, 3:4 and 21:9. If you start from your own image or clip, the video takes that file's shape.
Is there a watermark?
No. Videos are delivered clean on every option.
Can I use the videos commercially?
Yes. You can use what you make in ads, on social media, in products and for clients. The terms set out the details.
Which AI models does it use?
You choose in the studio: Wan 3.0 for the best value, Seedance 2.0 Mini for the lowest cost, or Seedance 2.5 for the most detail. All of them make video with sound.
Can AI video put text or my logo on screen?
Written words inside generated video are unreliable with every current model, so letters can come out wrong. Leave signs blank and add titles or a logo afterwards in any free editor.
What happens if a video fails?
You are not charged. If a render fails or is cancelled, the credits go back to your balance automatically.
Does it work on a phone?
Yes. It runs in any modern browser on phones, tablets and computers, with nothing to install.
Your first video starts with one sentence
Sign in, describe your idea and see it on screen. No card needed to try it.
Make a video free