AI-Powered Optimization

Your images.
More possibilities.

Nine focused AI capabilities. One familiar image pipeline. Choose a tool, review the result and put it to work.

Portrait source image used to illustrate the AI workflowYOUR ORIGINAL STAYS YOURS
An intentional workflow

Create with AI.
Publish with confidence.

AI suggests. Your team decides. Every generated image or video starts as a private preview, with an explicit approval step before public delivery.

Choose an image from a published source
Run a tool and inspect the result
Approve, copy the URL and reuse
Workflow illustration. The frame shown is illustrative, not a live detection result.
Nine capabilities. Real work.

From a source image
to something ready to use.

01
Smart crop

Make room for the subject.

Turn an original into framing suggestions for squares, wide banners and portrait layouts. Review the suggested crop before it becomes a delivery image.

Subject-aware framing · Four layout presets
02
Background removal

Keep what matters. Lose the backdrop.

Create a transparent PNG for products, people and editorial cards. Inspect fine edges against the checkerboard preview, then approve the cutout for delivery.

Transparent PNG · Local AI extraction
03
Alt text & tags

Give your images a voice.

Generate concise image descriptions and useful tags in Hebrew or English. Review the suggestion in the context of your page before using it as alt text.

Hebrew & English · Copy-ready suggestions
04
Privacy blur

A little more privacy.

Detect faces and vehicle license plates, then inspect a proposed blur before publishing. Detection may miss details; your team reviews the full image.

Faces & plates · Review before publishing
05
Enhance & upscale

Give small images more room.

Create 2× or 4× output with AI reconstruction and optional noise preprocessing. Review faces, text and fine details: generated detail is an estimate.

2× / 4× · Generative enhancement
06
Generative fill & expand

Make space for the story.

Generate missing edges for banners, cards or stories. Choose exact canvas dimensions or a layout preset, position the original and describe the surrounding background. Review the generated area before delivery.

Exact canvas sizes · Protected source area
07
Natural-language search

Describe it. Find it.

Search your indexed workspace images in Hebrew or English. Add images from published sources in background batches, then find them with a phrase instead of a filename.

Private workspace index · Up to 1,000 images
08
Image to video

A still image. A little motion.

Generate a short clip from a source image with a gentle zoom, pan or custom motion direction. Watch the complete preview before approving delivery.

4 / 6 / 8 seconds · 720p · No audio
09
Background replacement

Same subject. A different world.

Extract the foreground and create a new background from your prompt. Try a modern office or describe your own scene, then review the edges and lighting.

Local foreground matting · AI-generated scene
Built into your workflow

Generate once.
Keep the possibilities.

Approved crops and cutouts become saved image assets. The existing URL API handles their output sizes and formats.

Explore the AI API

Reuse the result

Saved results are reused for the same source image, tool and settings during retention. Another output size does not require another AI call.

A separate AI budget

Workspace operation limits keep provider attempts separate from standard image transformations and delivery. Provider fees are additional; AI pricing is quoted separately.

Review before publishing

Keep the original. Preview the generated image. Approve it before creating a public delivery URL. Descriptions remain suggestions for your team to review.

A considered rollout.

AI Studio is a feature for configured workspaces. Smart crop and descriptions require AI processing to be enabled. Background removal uses the configured local foreground model; a connected remove.bg account is optional. Image-to-video requires video generation to be enabled, model access and a separate video attempt limit; generated previews may incur provider charges before approval. Availability depends on provider access and workspace limits. Image-to-video also requires video generation to be enabled and has a separate attempt limit. Saved images and videos expire according to the source retention policy.

AI can miss a subject, leave imperfect edges or suggest an inaccurate description. Review your own images and intended layouts before publishing.

On the horizon

More room to create.

OCR is a possible addition for a future release. Background replacement, removal, expansion and enhancement are included in the configured AI Studio. Generative expansion and enhancement are included in the configured AI Studio.

Big on images.
Light on everything else.

Let’s build a simpler image pipeline for your next million requests.

Enterprise image optimization. Built for scale.