Create, edit and publish, then run your website and your app from their own dashboard. Every feature shown in real screenshots.
See the features
A Next.js project in React and TypeScript, sent to your GitHub repository. A spec change made in the repository comes back to Eqlo.
See the developer page
Paste a URL. In thirty seconds you see exactly what Google receives before any JavaScript runs.
Check a URLIn front of an image, everyone gives an opinion. In front of a screen that responds, you see where people get stuck. To turn a Figma design into a working app, attach the screen export to your request: Eqlo rebuilds its structure, makes it clickable and applies your exact values. You try the interaction yourself, without waiting for engineering to have a free slot.
BEFORE · drawn screen · AFTER · clickable screen
You worked out the reading order, the contrast and the rhythm. In a meeting, those choices become one opinion among many, because nobody can check them in use. Once the screen responds to clicks, the debate changes: people watch where users get lost.
Comments arrive by chat, by email or in a spreadsheet, far from the element they refer to. You then have to guess which button each one is about. A comment pinned to the section in question settles that.
A sequence that does not work often shows up once it has been built, when fixing it costs the most. Trying the interaction before development flips the order: you fix things while it is still free, and engineering receives a screen that has already been tried.
A file, a page of notes, a folder of exports: the developer fills in the gaps, gets a spacing wrong or invents an empty state. The gap appears at review. A screen that already works, with its values written down plainly, leaves far less room for doubt.
This question always comes up, and it deserves a precise answer. Export the screen as an image from Figma, Sketch or Penpot, then attach it to the request with the paperclip, by drag and drop, or by pasting. Up to six files of 15 MB each fit in one request.
What the AI rebuilds. The structure of the screen: the zones, the order of the blocks and the columns. The text hierarchy, from heading to subheading, label and data. Common interface elements: cards, fields, buttons, lists, tabs. The result is a clickable screen, not one more picture.
What is better written down. The hex codes of your colors and the names of your fonts: written in the request, they are applied as is, whereas an image only gives an approximation. A corner radius in pixels is brought to the nearest of the five levels the engine offers. The states nobody drew, such as an empty list, loading or an error. And the rules of the flow, which live in your notes, never in the image.
What the image does not carry. Component variants, auto-layout constraints and prototype links stay in your design tool. The AI works from what is visible on screen. That is a real limit, but it has an upside: any design tool will do.
A design system wears down as soon as each value gets reinterpreted. Write them once, and they hold for the whole project.
Where to write them. In the current request or, better, in the project's permanent instructions: the AI rereads those 2,000 characters with every new request. Your values then stay in place as the project grows.
After that. The Design tab changes the style, palette, fonts, animation level and the layout of each section without AI, so at no cost. Corner radius, shadows or button shape take one sentence in the chat. A locked section is no longer touched by the AI, even during a full redesign. Up to sixty sections can be locked.
Three concrete results, with times given as a rough guide. None of them goes through the development queue.
The attached image becomes a screen that responds: you type into fields, filter a list, follow a button to the next screen. You ask for missing states in one sentence. The preview shows a desktop view or a 390 × 844 phone view.
The colors and fonts you wrote down apply to every page. To weigh two options, make both requests, then compare the two versions side by side. No version is ever deleted.
The project goes to your GitHub repository with its specification and the complete code in Next.js, React and TypeScript. Whatever a developer changes in the repository comes back into Eqlo as a new version. Nobody starts again from a screenshot measured pixel by pixel.
Several accessibility points are handled by default on every published site. They spare you the most repetitive fixes, without replacing a check.
What is in place. A "Skip to content" link at the top of the page and an accessible mobile menu. Contrast corrected automatically to reach the AA level. A description written by the AI for every image, editable afterward. And animations switched off when visitors have asked their device for reduced motion.
What to check yourself. Whether image descriptions that carry meaning are accurate, since you know them better than anyone. The full keyboard path, which takes five minutes to test. And the clarity of labels, especially on action buttons.
What no tool does. Formal compliance is proven by an audit. Eqlo improves the starting point, but it does not issue any certificate.
A scenario invented for illustration. A designer works on the app for a group of yoga studios. She has drawn five screens for booking a session, and a review is set for Friday with the product lead and two developers.
On Wednesday morning, she attaches the five screens to her request and writes down her primary color, her font and her corner radius. A few minutes later, the screen lists the classes with their length, room and price, the next times and the Continue button, in her color and font. Two gaps stand out at once: nothing was planned for a week with no free session, and the confirmation screen had no back button.
On Thursday, she sends the review link to five colleagues, who pin their comments on the sections they mean. She passes the comments to the AI in one click. On Friday, everyone has already used the flow, and the review ends in a decision. The developers then pull the code into their repository.
Figma, Sketch or Penpot remain where you draw and explore. Eqlo comes in when the screen needs to become usable. Both tools fit into the same working day.
Your color codes, fonts and measurements are written once in the project's permanent instructions. The AI sticks to them instead of inventing a palette. The consistency work your team has already done pays off directly.
Same day, same people. The difference is what gets shown: a flow everyone has already gone through, with comments pinned where they belong instead of a thread to untangle.
Making a screen usable, getting people to try it and adjusting it needs nobody else. That is the stage where a designer's eye matters most, and the one where each round trip with developers wastes the most time.
When the flow needs to become a real product, the complete code goes to your GitHub repository, and changes made there come back into Eqlo. Developers keep the screen layout and your brand rules as a base. They then write the business logic and access rights according to their own standards.
Where to go next, depending on your role in the project.
No. An exported image of the screen is enough, attached to the request. Because the AI starts from what is visible, it does not matter which tool you drew it in. Add your exact values in writing, since an image only gives them roughly.
Yes. Attach up to six images to one request and state the order of the flow. For a longer flow, add the next screens in a new request once the first ones are in place.
Yes, if you write them down. Hex codes and catalog fonts given in the request are applied as is, and a corner radius in pixels is brought to the nearest level. Only contrast that is too weak gets corrected to reach the AA level.
They stay in your design tool. The AI builds a layout that adapts to the visitor's screen from the image. If the result does not suit you, point to the section in the preview and describe the change.
The generated screen often includes some, and that is the most common gap with the mockup, which rarely draws them. You adjust them by writing your own wording, or by editing the text directly in the preview.
No. Reviewers open a private link, or scan its QR code on a phone, then type their first name, pick a section and leave a comment of up to 1,000 characters.
Yes. The complete Next.js, React and TypeScript code goes to their GitHub repository, and their changes come back into Eqlo. They usually keep the screen layout and the brand rules, then rewrite the business logic.
Five projects come free, and no card is asked for. Your reviewers comment with nothing more than a first name.