Lexica

Lexica developed an AI image search and generation platform that brings visual discovery and image creation into one interface. Users could browse images and prompts, learn from examples, and generate their own images with Lexica Aperture, its text-to-image model family. Its documented Search API lets developers search by text or supply an image URL for reverse image search, returning 50 image results in JSON.

Founded by Sharif Shameem, Lexica began in 2022 as a searchable collection of Stable Diffusion images and prompts. Its search evolved from matching prompt text to semantic retrieval using CLIP image embeddings and FAISS, allowing searches to reflect image content. For generation, the team used visitor preferences to train aesthetic scoring models and filter training data. In 2023, Shameem reported over five billion images served per month; he now works on OpenAI’s Labs team.

Lexica’s published licensing connects its creative tools to personal and business use: free plans permit personal use, while paid plans permit commercial use of images, including those created by other users. Pro and Max provide company licenses, with private generation reserved for Max. Commercial rights to images downloaded or generated during a paid subscription continue after cancellation.