how to make an ai video from photos

How to Make an AI Video From Photos: A Complete Walkthrough

From two photos to a finished clip, step by step.

9 min read

Reviewed against the live product on 2026-10-05. The generator offers 480p or 720p, one photo per performer, and a template-set length.

Try the generator
How to Make an AI Video From Photos: A Complete Walkthrough

Making an AI video from photos is a short process, but a few choices early on decide how good the result looks. This walkthrough covers the whole path: what to prepare, how to choose images, how to upload and assign performers, how to pick a template, how to generate, and how to get the finished file.

It is written for someone doing this for the first time. There is no prompt to write, no timeline to edit, and no software to install — the generator runs in the browser and the output is a standard video file. The main skill involved is choosing the right photos.

Read it end to end once, or jump to the step you are on. Either way, the goal is the same: a clip you are happy to keep, produced without any of the usual video production overhead.

Make your video here

Choose one photo per performer, preview them and swap positions. Generation requires sign-in and uses your account credits. Use photos you have permission to upload.

Choose your performers

0/2
Left · Photo A
Right · Photo B

Clear front-facing photos work best — one person per image, no covered faces.

Classic Duo

Two people performing a short studio duet, recast with your two photos.

Available
ResolutionNo duration selector — set by template

Each generation uses credits from your balance.

Sign in to generate videos.

Signing in keeps selected photos in this browser for your return. The draft expires after 20 minutes and is removed on return or the next visit after expiry.

Template preview — your two photos replace these performers.

Preview and change the motion template
Available

Classic Duo

Two people performing a short studio duet, recast with your two photos.

10s2 people
Available

High Energy Duo

A faster, more upbeat two-person studio performance.

15s2 people
Available

Soft Groove Duo

A slower, relaxed two-person studio groove.

15s2 people
Available

Still to Motion

A still portrait brought into gentle motion.

10s2 people
Available

Minimal Motion

Subtle movement with a minimal set and camera.

13s2 people
Available

Kangaroo Boxer

A single boxing-kangaroo motion source from your provided material.

15s1 person
Available

Duo Conversation

Two performers facing each other in a dialogue scene.

10s2 people
Available

Cinematic Duo

A wider, more cinematic two-person framing.

15s2 people

What you need before you start

The list is short. You need two photos — one per performer — a moment to pick a template, and a little patience while the render runs. That is the entire setup. There is no camera, no microphone, and no editing application involved.

The photos are the part worth preparing. JPG, PNG, and WebP files are accepted, up to 10 MB each. Within those formats, the quality of the image matters far more than the format itself: a clean, well-lit portrait will outperform a large but blurry one every time.

It also helps to know what you want before you upload. Deciding whether you want a lively clip or a calm one makes template selection quick, and it saves you from generating a result that works technically but misses the mood you had in mind.

  • Two photos, one for each performer.
  • JPG, PNG, or WebP, up to 10 MB per image.
  • A rough idea of the mood you want, calm or energetic.
  • A few minutes for the generation to complete.

Step 1: choose the right photos

Start with subject clarity. Each photo should have one obvious person in it, positioned so the face is easy to read. A portrait taken straight on, or close to it, gives the model the cleanest view of the face. Strong angles and profiles give it less to work with.

Then check the light. Soft, even lighting across the face transfers best. Harsh shadows, backlighting, or a strong colour cast all create conditions the model has to compensate for, and those compensations can make the final clip look less like the original person.

Finally, check what is covering the face. Sunglasses, masks, or a hand in front of the face will reduce how much detail the model can carry into the scene. A hat is usually fine as long as the face remains fully visible. If you have a choice between two photos, pick the one with less obstruction.

Step 2: upload and assign the performers

Open the generator at the #generator anchor on the home page. You will see two photo slots. Upload one image into each: the left slot takes one performer, the right slot takes the other. The tool treats them as two separate identities and keeps them distinct throughout the scene.

If the order matters to you — for example, if one person should appear on a particular side — check the arrangement before you continue. Use Swap to exchange the two slots, which flips which photo fills which performer role. Getting this right before generating is much easier than fixing it afterwards, because changing the order means a fresh run.

If you only have one usable photo, that is fine — but choose a single-performer template instead of a duet one. Uploading one photo into a scene staged for two people will not produce a pair.

Step 3: pick a preset template

The template is your creative choice, and it determines the scene, the movement, and the length of the clip. Eight presets are available, and they range from energetic two-person performances to gentle single-portrait motion.

Match the template to the mood of your photos. Bright, playful images suit High Energy Duo. Calmer portraits pair well with Soft Groove Duo or Minimal Motion. If you want a more composed, film-like framing, Cinematic Duo widens the shot, while Duo Conversation places the two performers facing each other.

If you are uncertain, Classic Duo is a sound default for a pair of portraits, and Still to Motion is the restrained option if you want subtle movement rather than a full performance.

  • Classic Duo — balanced two-person studio duet, a safe default.
  • High Energy Duo — fast and upbeat, for lively photos.
  • Soft Groove Duo — slower and relaxed, for calm photos.
  • Still to Motion and Minimal Motion — subtle, restrained movement.
  • Kangaroo Boxer — single performer, for when you have one photo.
  • Duo Conversation and Cinematic Duo — composed, film-like two-person framing.

Step 4: choose a resolution and generate

Two resolutions are available: 480p and 720p. 480p generates faster and produces a lighter file, which makes it ideal for testing. 720p takes longer and yields a sharper result, which is worth it for the version you intend to keep.

A practical routine is to run your first attempt at 480p. If you are comparing templates, this lets you see your own faces in each scene quickly and decide which fits. Once you have a favorite, re-run it at 720p for the final file.

There is no separate duration setting — the template fixes the length. Once you start the generation, the status indicator tracks progress. Video generation is heavier than image generation, so give it time rather than reloading the page mid-run.

Step 5: preview and download

When the render finishes, preview the clip in the page. This is the moment to check the two things that matter most: whether each performer resembles the source photo, and whether the overall mood matches what you wanted.

If the likeness is off, the photo is usually the cause. Replace the weaker upload with a clearer, more front-facing image and generate again. If the mood is off, try a different template instead — the same faces will read very differently in a fast scene than in a slow one.

When you are satisfied, download the MP4. It is a standard video file, so it plays in any normal video app and can be shared or edited with the tools you already use.

Troubleshooting checklist

Most disappointing results trace back to one of a few causes, and each has a straightforward fix.

  • The person looks less like the photo: use a clearer, front-facing portrait with even lighting.
  • One performer looks out of place: match the lighting and framing of the two uploads more closely.
  • The clip feels too still: switch to a higher-energy template such as High Energy Duo.
  • The clip feels too busy: switch to Soft Groove Duo, Minimal Motion, or Still to Motion.
  • Only one person appears: you uploaded one photo into a two-person template — add the second photo.
  • The performers are on the wrong sides: use Swap before generating, then re-run.

Getting better results over time

The single biggest improvement comes from your photo library. Keeping a few good, front-facing portraits of the people you generate most often means you always have a strong input on hand, rather than hunting for one under time pressure.

The second is familiarity with the templates. Once you have seen how your own photos behave in a fast scene versus a slow one, you will be able to predict which template suits a given pair without testing. That instinct develops quickly, because the scenes are stable and repeatable.

The third is patience. Generation takes longer than editing a photo, and a failed or cancelled run returns its credits to your account automatically, so there is no penalty for simply trying again. Treat the first attempt as a draft and the process stays relaxed.

Frequently asked questions

How many photos do I need to make an AI video?

Two for a duet template, one photo per performer. A single-photo template such as Kangaroo Boxer works with one image.

Is a text prompt required?

No. You choose a preset template instead, and the template defines the scene, the movement, and the clip length.

What photo formats and sizes are supported?

JPG, PNG, and WebP up to 10 MB each.

How long does generation take?

Generation takes a few minutes and the page tracks its status. 480p runs finish faster than 720p, so you can test quickly before committing to a higher resolution.

What do I get at the end?

A downloadable MP4 video file that plays in any standard video app.

What if the result does not look like the photos?

Replace the weaker photo with a clearer, front-facing portrait that has even lighting, then regenerate. Photo quality is the main factor in how closely the clip resembles the original.

Turn your photos into a duo video

Upload one photo per performer, pick a preset template, and generate a clip you can download. No prompt and no editing required.

Open the generator