Turn any image into a branded, watchable video in minutes.
Upload screenshots, product photos, or design exports. ngram reads each frame, adds context-aware motion and transitions, layers in captions and voiceover, and exports the same set in landscape, square, and vertical.
Drop an image or a folder, or click to browse
JPG, PNG, WebP, TIFF, SVG, HEIC · up to 20 MB per image, up to 50 per project
Trusted by teams at
How it works
Four steps. Static images out, watchable video back.
No timeline, no After Effects keyframes, no slideshow templates. Drop the images, accept the AI-built storyboard, ship the video.
Drop the images in
One image, a folder, or a Figma export. We accept JPG, PNG, WebP, TIFF, SVG, and HEIC. Order them on upload or rearrange later.
ngram reads every frame
Each image is classified — UI screenshot, product photo, diagram, illustration — and gets motion picked to fit: smart push-in on UI, parallax on photos, sequential reveal on diagrams.
Tune the storyboard
Reorder scenes, drop in voiceover from a written script, tweak per-scene duration, swap the transition. Brand Kit fonts, colors, and intro card apply on the fly.
Export every aspect ratio
One render produces 16:9, 1:1, and 9:16 — captions reflowed, focal point preserved per ratio. Download MP4, GIF, or WebM, or post straight to LinkedIn, X, or YouTube.
Output controls
Smart defaults for the static-to-motion jump. Real knobs underneath.
Per-image motion choice
Smart push-in for screenshots that have a focal button. Parallax for product shots with depth. Hold-and-pan for wide hero images. Override any pick per scene.
Photomotion on stills
Cinematic camera moves — push-in, parallax, slow pan — applied frame-by-frame so a flat photo reads like footage. Useful when source material is image-only.
Captions matched to the image
Auto-captioned from the voiceover or script. Caption position dodges UI elements detected in screenshots, so no text covers the button you wanted to show.
16:9, 1:1, 9:16 from one set
Same image set, three ratios in a single render. We reframe per format and re-center on the detected focal point so nothing important crops off-screen.
Brand Kit applied automatically
Logo lock-up, color treatments, font choices, intro and outro cards pulled from your Brand Kit. Every export looks like the same team made it.
Voiceover from a written script
Type the narration per scene; ngram generates AI voiceover (ElevenLabs or MiniMax) timed to image transitions. Choose a voice once, reuse across videos.
Batch a folder of products
Drop a folder of product shots, get a video per SKU in parallel. Same Brand Kit, same script template, different image set per render.
Source files gone in 24h
Uploaded images and renders auto-delete after 24 hours. Processed in-region, never used to train models. in-region processing.
The rest of ngram
Converting the image is one step. The product is everything around it.
AI Visuals
When your image set is missing a transition shot, AI Visuals generates a scene-matched fill in your brand's style so the cut from screenshot to product photo does not jar.
Learn moreMotion Graphics
Animated callouts, lower-thirds, and labels layered over each image. A 'New' badge on a UI screenshot or a feature label on a product photo without leaving the editor.
Learn moreAI Voiceover
Type a script per scene; voiceover gets generated and timed to image transitions. Useful when your stills carry the visual story but need narration to land.
Learn moreCaptions
Burned-in captions that route around detected UI elements in screenshots, so the call-to-action button is never covered by subtitle text.
Learn moreBrand Kit
Logo, colors, fonts, intro and outro cards applied automatically to every image-driven video. Ship 10 product videos that all look like the same brand.
Learn moreMulti-format Export
Render 16:9, 1:1, and 9:16 from one image project. Each ratio gets its own reframe and focal-point lock so the hero shot stays in frame on any platform.
Learn moreUse cases
Where image-driven videos earn their keep.
Launch a feature with screenshots, not a demo recording
PMM has the new-feature screenshots before there's a live demo. Stitch them with smart push-ins and a voiceover script and the launch video is ready before the build is.
See use caseChangelog clips from UI screenshots
Every release ships a row of screenshots in the doc. Turn that row into a 30-second changelog video for the email and the in-app banner.
See use caseProduct videos from product photos
Drop a folder of product shots, get a video per SKU. Captions, pricing overlay, and brand colors applied per scene — sized for feed and Reels.
See use caseVideo for the campaign email hero
Take the static email hero image, give it 6 seconds of motion, drop it in as a GIF or inline video. Open rates climb without a new asset request.
See use caseLanding page hero from design exports
Designers ship Figma exports. Animate them into a 10-second hero loop above the fold instead of a static image, without booking video production.
See use caseAPI doc walkthroughs from screenshots
Code screenshots and dashboard screenshots become a moving walkthrough on the docs page. Devs see the click, the response, and the next step in sequence.
See use caseOnboarding videos from UI screenshots
Pair the existing in-app screenshot set with voiceover and you've replaced the 'getting started' help article with a 90-second video.
See use caseStatic ad creative gets a motion variant
The ad-team's static set becomes a motion variant for Meta and TikTok in the same render. Same brand, same focal point, three ratios.
See use caseHelp articles that show the click
Screenshots inside the help article become a 20-second clip at the top of the page. Fewer support tickets per article, same source material.
See use caseOther converters
Starting from a different file type? There's a converter for that.
Image to video is one of 16+ ngram converters. They all share the same engine, the same Brand Kit, and the same storyboard control surface.
Image to video, but tuned for sequences of product UI captures. Smart zoom finds the button you meant to highlight.
Open converterIf a single screen recording is closer to what you want than a stitched image sequence, start here. Cursor smoothing, dead-air trim, branded chrome.
Open converterDecks, release notes, or product one-pagers as the source. ngram extracts each page as a scene and animates from there.
Open converterTools that pair with this converter
Refine the image set. Edit the output.
Editing further
After the image-to-video render lands
Video Editor
Open the rendered video on a timeline and adjust per-image duration, swap a screenshot, or extend the hero hold without re-uploading.
Open toolAdd Subtitles to Video
Burned-in or.srt subtitles layered onto the image-to-video output. Position them away from the UI region each screenshot wanted to highlight.
Open toolVideo Cutter
Trim the finished image-to-video render into shorter cuts — a 6-second teaser for ads, a 30-second cut for LinkedIn.
Open toolAdd Music to Video
Layer licensed background music under the voiceover and image transitions. Auto-ducks under narration so the script stays clear.
Open toolGenerating from scratch
When you do not have source images yet
AI Image Generator
Generate the missing scene image, then feed it back into image-to-video as another frame. Useful when you're one shot short of a full sequence.
Open toolAI Video Generator
If you do not have images at all, prompt a full video. Comes back as scenes you can later mix with your real image sequence.
Open toolAI Avatar Video Generator
Pair the image-to-video output with an AI avatar intro that narrates the sequence. Same Brand Kit, one render.
Open toolText to Speech Video
Write the voiceover; get a narrated clip with stock or generated visuals you can splice into the image set.
Open toolPolishing the source
Fix the images before they hit motion
Background Remover
Strip the background from a product photo before it enters the storyboard — the parallax move reads cleaner on a transparent subject.
Open toolVideo Background Remover
Use after rendering if you want the image-to-video subject isolated for compositing into another brand template.
Open toolHero image generator
Generate a missing hero shot in the same style as your existing screenshots, then drop it into the image-to-video sequence as scene one.
Open toolVideo Compressor
Compress the final render down for email or in-app embed without losing the cinematic motion applied per image.
Open toolBuilt for teams
Who reaches for image-to-video in your company?
Product Marketing
Launch videos built from feature screenshots before the live demo is shipped. Same source images, three ratios, one Brand Kit.
See workflowsEcommerce
Product photos become a video per SKU. Batch the catalog overnight; ship feed-ready 1:1 and Reels-ready 9:16 in one pass.
See workflowsGrowth Marketing
Static ad creative gets a motion variant in the same render. Same focal point, same brand, video unit ready for Meta and TikTok.
See workflowsProduct Managers
Turn the screenshots already in your release doc into a changelog clip without booking time with marketing.
See workflowsDeveloper Relations
Dashboard and code screenshots become an integration walkthrough that lives in the docs and on YouTube.
See workflowsSupport Teams
Help-article screenshots become a 20-second clip at the top of the page. Less ticket volume, no new asset request.
See workflowsContent Creators
Photo dumps and design exports become Reels-ready vertical video with captions and music. One render, all three ratios.
See workflowsSales Enablement
Reps need a 30-second product clip and only have screenshots. Image-to-video stitches them with voiceover and ships before the call.
See workflowsIntegrations
Trigger image-to-video from where your assets already live.
Each integration ships with a working recipe. Start from one, or wire your own with the REST API and webhooks.
whenA new image lands in a Drive folder
thenRun image-to-video with the team's Brand Kit and post the result to #launches
whenAn agent calls the image-to-video tool with a list of image URLs
thenReturn a rendered MP4 and a hosted share link
whenYour S3 bucket gets a new product-photo batch
thenConvert each image set to video; no asset leaves your VPC
whenA new product is published in Shopify with images attached
thenConvert the product images to a video ad and push to Meta and TikTok
whenYou right-click a Figma export or a product photo in the browser
thenSend the image set to image-to-video and get the render in a new tab
whenAn image-to-video render finishes in 1:1
thenSchedule the post on the company page with the caption already drafted
whenAn image-to-video render finishes in 16:9
thenUpload to the channel with title, description, and end card pre-filled
whenAn image-to-video render finishes in 1:1 or 9:16
thenPost to X with copy and the rendered short attached
How it compares
If you've been using another tool for this.
Canva and Kapwing stitch slideshows. Runway and Pika invent footage from a prompt. ngram keeps your real images and adds motion, voiceover, captions, and branding around them.
| Feature | ngram | Canva | Kapwing | Runway |
|---|---|---|---|---|
| Keeps your original images intact | Yes — motion is layered over real files | Yes | Yes | No — generates new footage |
| Per-image motion choice | AI picks push-in, parallax, or hold-and-pan per frame | One template motion for every slide | Manual keyframes per scene | Prompted from scratch |
| AI voiceover from a written script | ElevenLabs and MiniMax, multilingual | Limited TTS voices | Limited TTS voices | — |
| Captions that dodge UI elements | Caption position aware of detected UI regions | Manual placement | Manual placement | — |
| Brand Kit applied automatically | Logo, colors, fonts, intro and outro per render | Manual setup per project | Manual setup per project | — |
| Multi-format export from one project | 16:9, 1:1, 9:16 in one render | Duplicate and re-edit | Resize after | Single ratio per generation |
| Batch a folder of images | Parallel renders, one video per SKU | One project at a time | Sequential queue | Single generation |
| API and webhooks | REST, MCP, n8n, Zapier | Limited API | API on enterprise | API available |
| Source files auto-deleted | 24h, in-region, never trains models | Variable | Variable | Variable |
FAQ
Common questions about image to video
Still curious?
Image → Video
Turn the images you already have into watchable video.
Drop one image or fifty. ngram reads the frames, picks the motion, layers the voiceover, and ships landscape, square, and vertical in one render.