
Version 1 of Ailix.ai came out in late February 2025. By then, the catalog had grown from thirty-five to nearly sixty models. Claude 3.7 Sonnet, GPT-4.5 and Gemini 2.0 Flash-Lite were added within days of their release, which is what an aggregator is for and hardly an event.
The real change was elsewhere. The beta feedback all said the same thing: knowing that a model exists is not enough to use it. Faced with a blank page and a blinking cursor, an employee does not know what to type.
So V1 added two things the catalog did not provide. Assistants are created in a few clicks for a specific task (drafting a proposal, analyzing a dataset, producing a visual) and keep their instructions from one use to the next, so nobody rewrites the same request every Monday. Agents chain several steps without being prompted again at each one.
Video generation from text or an image arrived with this version, along with image upscaling. Marketing teams that produced their visuals elsewhere stopped leaving the platform to do it.