Until now, images in AYETO lived in the chat: you asked for a picture, got one back and downloaded it. That works for a quick idea, but real work on an image means more than that. You want to try a few styles, fix the light, remove something that should not be there, add a headline and keep everything organised for later.

Image Studio is where that work happens now. Open it from the left menu. It has three parts, switched at the top of the screen:

Generate: create new images from a description, optionally in one of 61 artistic styles.
Editor: a full photo editor with free manual tools and paid AI tools.
Library: every image you generated, uploaded, edited or saved from chat, in folders.

Manual editing is free. Only generating and the AI tools use credits, and you always see what a result cost.

Here is how it works on one example: a social media post for a café's autumn menu.

Generate: from a sentence to a picture

Describe what the image should show. We wrote: "A ceramic cup of cappuccino with latte art on a wooden table in a cozy café, rain on the window behind it, soft morning light, autumn leaves outside."

Then pick a style. There are 61 of them, from Photorealistic, Watercolor, Pixel Art and Isometric 3D to painters like Monet, Hokusai or Klimt. Each tile shows a preview, so you see the style before you pay for it. Choose a model (the default GPT-Image-2.5 Flare suits everyday work, Nano Banana Pro and the others are one click away), the size and how many images you want. Press Generate.

Generate view with a café prompt, Photorealistic style and two generated images

Every result has Open in editor and Variant, which uses the image as a starting point for a new version.

Not sure what the style does to your prompt? Preview prompt shows the exact text the model will get, without spending anything. Below you can see the same prompt in the Watercolor style: the style adds its own description and a list of things to avoid.

Prompt preview for the Watercolor style above two watercolour versions of the café picture

A few more things to help with generation:

References: add up to four pictures (from the library or your computer) for the model to start from, such as your product, your logo or a colour palette.
Prompt templates: ready-made prompts for logos, business cards, product shots, interiors, events, social media and thumbnails. Pick one and replace the details with your own.

Library: all your images in one place

Everything you generate lands in the Library straight away. So do uploaded photos, edited versions and images you save from chat. Filter by origin (Uploaded, Generated, Edited, Chat, Favorites), search by name, and sort images into folders. Select several at once to move them, delete them or use them as references for the next generation.

Library with the Café campaign folder with the generated café images

Double-click an image, or press Open, to edit it.

Editor: free manual editing

The editor has the tools you know from photo apps, arranged in a rail on the left:

Crop: aspect ratios, straightening, rotation, flipping and keystone correction.
Adjust: light, colour, detail, effects, colour grading, lens, a tone curve and an HSL mixer.
Looks: one-click presets, including film emulations (Kodak Gold, Portra 400, Fuji Velvia, CineStill, Tri-X…). You can import your own .cube LUT and save your settings as My looks for later.
Masks: linear, radial, brush, luminance and colour masks for local adjustments.
Retouch: heal and clone brushes.
Text: headlines, shapes and ready-made layouts.

Editor with the Looks panel open and the Kodak Gold look applied to the café photo

Press the Before/After button at the bottom right (or hold Space) to compare with the original at any time. Nothing here costs credits.

AI suggest: let the AI set the sliders

Not sure where to start? Open AI suggest, optionally say what you want, and press Ask the AI. A vision model looks at the photo and proposes concrete slider values, and it can even add a local mask. You see the result live before anything is committed. The Strength slider tones the proposal down, Apply keeps it and Discard throws it away.

We asked for "Cosy autumn mood, warmer tones, draw the eye to the cup". The AI kept the warmth, calmed the bright window and added a gentle local lift on the cup:

AI suggest panel with the explanation and the proposed changes, strength slider and Apply button

Besides Suggest settings there are two more modes: Copy the look turns the grade of a picture into a reusable look, and Critique tells you what is wrong with the photo and which sliders would fix it. A proposal you like can be saved with Save as look.

AI tools: change what is in the picture

Sliders cannot add a croissant to the table. AI tools can. They are grouped as:

Enhance: auto enhance, upscaling, super resolution, detail enhance, face enhance, noise removal.
Background: remove or blur the background, replace the sky.
Retouch: remove or replace objects, age transform.
Creative: image variation, sketch to image, colour harmonisation, relighting, pose guide.
Restore and Geometry: colorize, perspective fix, outpainting, image expand, depth map.

We used Object Replace: "the candle in the glass holder on the left" → "a golden butter croissant on a small ceramic plate". Every AI result opens in a before/after preview with its price. Apply makes it the new original, and your manual edits carry on on top of it. Discard throws it away.

Before and after preview of Object Replace: the candle is replaced by a croissant on a plate

Changed your mind later? Revert AI in the toolbar goes back to the picture before the AI step.

AI styles apply any of the 61 styles to a photo you already have, so the same café shot can become a watercolour, an anime frame or a Van Gogh.

Text: from photo to post

The Text stop turns the image into a finished graphic. Add text and shapes, style them (font, weight, colour, background, spacing, rotation), or start from one of the Layouts: lower third, centred poster, quote card, sale banner, event card, title band and more. Every element is a layer you can move, lock, hide or reorder.

Text panel with the title band layout and the headline “The autumn menu is here” on the café photo

History, saving and export

Every step is listed in History, and you can jump back to any of them. The history is saved with the image, so when you open it next week you can still undo.

History panel listing Original, Look Kodak Gold, AI Develop, Object Replace, Apply layout and Edit layer

Save to library stores the edited version next to the original. Export downloads it as PNG, JPEG, WebP or PDF, in the original size or scaled down, optionally with a watermark.

Image Studio and chat

Image Studio is connected to the chat in both directions:

In the chat input, the paperclip now offers Pick from Image Studio. Choose up to ten pictures from your library and ask the assistant about them.
Images in chat (attachments and generated pictures) have a Save to Image Studio button, so a picture from a conversation can go straight into the editor.

To finish our example, we picked the finished post from the library and asked the assistant for an Instagram caption:

Chat with the finished café image attached and an Instagram caption written by the assistant

Organisations and credits

If you work in an organisation, Image Studio follows the organisation you have selected: images, folders and looks are kept per organisation, and generation and AI tools are charged to the organisation's credits. Each result in the library shows what it cost. For an edited image, that includes the AI steps it went through.

Try it

Open Image Studio from the left menu, go to Generate and describe the first picture that comes to mind. Or drop a photo you already have into the editor. Manual editing is free, so there is nothing to lose.