Diiverge

Diiverge is a web platform that turns a single picture into a playable, point-and-click adventure. It presents itself as a series of explorable worlds, where each volume is grown from one seeded image and becomes a persistent adventure that visitors can walk through. A player clicks something in the frame, chooses what happens to it, and the system generates the next scene together with a short film of the moment in between. The public worlds are free to play, and the tagline sums up the promise: start with a picture and see where it goes. Diiverge is built for people who want to explore AI-generated stories rather than simply read them, and for creators who want to turn a photo, painting or screenshot into a world of their own. Conventional point-and-click adventures are authored in advance: artists paint the scenes, writers script the branches, and developers wire the choices together before anyone plays. The result is fixed, and every player walks the same limited set of paths. Diiverge takes a different starting point by treating a single image as the seed for an entire world. Instead of shipping a finished story, it grows one scene at a time as visitors interact with the picture in front of them. That approach also addresses a familiar limitation of AI-generated content, where images and clips tend to disappear the moment they are produced. On Diiverge, the scenes people carve out stay part of the volume, so the world accumulates structure and history rather than resetting to a blank slate. Playing a volume is deliberately simple. A volume begins as one seeded image, and glowing dots mark the things inside the frame that can be interacted with. You click one of those elements, choose what happens to it, and the platform responds by generating a new scene while the moment in between plays as a short film. From there you can keep going deeper into the world or go back and take a different route. Nothing about the interaction requires technical skill: the choices are offered on top of the picture, and the AI handles the artwork, the transition and the continuity. The set of available interactions is derived from the image itself, which is why any picture can serve as a starting point, whether it is a photograph, a painting or a screenshot. Persistence is the structural idea behind the whole series. Every path anyone takes is saved, so a scene that one visitor carves becomes part of the volume for good. A later visitor can walk the same path, and the volume's map shows every branch that has been taken so far. Shared links land on the exact scene they point to, which makes it possible to send someone directly to a moment inside a world rather than to its front door. Because each world grows as people explore it, the numbers attached to a volume are a record of that collective activity: The Crystal Pass is listed as Volume II with 2,831 scenes so far, while Volume I, The Lantern Cove, is listed with 3,033 scenes. The series is described as persistent point-and-click adventures, and that persistence is what separates them from one-off AI image or video generations. The studio is where people make their own worlds. You upload a picture, sign in with your email, and the studio turns that photo, painting or screenshot into an explorable adventure. You buy scenes in packs, and the adventure stays private until you choose to share it. Sharing produces a link that others can follow, and the paths they take are saved just like the paths in the public volumes. The economics are connected to the cost of generation: playing is free, and replaying scenes that others have already carved costs nothing at all, but generating a new scene costs real money. Because of that, each volume in the series comes with a fixed number of free scenes and stops growing once they are used up. Sponsors add more scenes, and those additions are for everyone. A sponsorship is bought as a tier starting from $25 and adds roughly two scenes per dollar to the volume chosen; the scenes never expire, and the sponsor's name goes on the volume's frames. Underneath the experience is a chain of specialised models, and Diiverge describes each step explicitly. A segmentation model cuts the things in the frame out and makes them clickable, which is what produces the glowing dots that mark interactive elements. A vision model decides what each of those things is and what it might do, translating the contents of the picture into plausible choices. When a player picks a choice, an image model paints the aftermath of that chosen event as the next frame, and a video model renders the moment between the two frames as film. Overseeing the sequence is a judge model that reads the story so far and cuts any choice that breaks its continuity. Together these steps mean a single still image can become a branching, narrated world with motion and a sense of consequence, without anyone hand-authoring the scenes. For players, the direct benefit is that the whole experience is free to enjoy and free to revisit. Replaying scenes that others have already carved needs nothing at all, so there is no cost to exploring a volume that has already grown. Every path is saved, which means the effort of exploration produces something permanent rather than a disposable result: the world you walked through is still there for the next visitor, and the branches you opened become part of the map. For creators, the studio provides a way to turn an ordinary picture into a shareable adventure without any art or engineering work, and the privacy default means a world can be built and reviewed before it is shown to anyone. Sponsors get their name placed on the frames of a volume and the knowledge that the scenes they fund are added for everyone and never expire. Concrete scenarios follow from those mechanics. A player can open The Crystal Pass, the latest volume, step in, and start clicking through its scenes, using the map to see which branches other visitors have already taken. Someone who wants to follow in another person's footsteps can open a shared link and land on the exact scene it points to, then continue from there. A creator who has a photograph, a painting or a screenshot they like can bring it into the studio, buy a pack of scenes, and shape it into a private adventure before sharing the link with friends or an audience. A sponsor can pick a volume and a tier, from $25, and add roughly two scenes per dollar so that every future visitor to that world has more to explore. Anyone who simply wants to look around can replay scenes that have already been generated at no cost. Diiverge runs on the web at diiverge.co and is presented as a game and world-building experience rather than a developer tool. The audience implied by its features spans casual players who want a free, interactive AI story, people who enjoy branching adventures and want to see where a choice leads, and creators who want to turn an image into an explorable world and share it by link. The pricing model is mixed: playing the public volumes and replaying existing scenes is free, while generating new scenes requires buying scene packs in the studio, and sponsorships that add scenes for everyone start at $25 per tier with roughly two scenes added per dollar. Worlds created in the studio remain private until their creator chooses to share them. The takeaway is that Diiverge treats a single picture as the starting point for something durable. Click something in the frame, choose what happens, watch the moment play as film, and leave behind a path that the next visitor can follow. Free to play, persistent by design, and open to anyone with an image to upload, it turns passive pictures into worlds that grow with every person who explores them.