Meta Platforms Inc. today debuted an image generation model that can write code and search the web. Muse Image is the second algorithm released to date by Meta Superintelligence Labs, the company’s artificial intelligence research group. The first is the Muse Spark large language model that made its debut in April. Both algorithms are available through the Meta AI chatbot. Muse Image can generate images based on highly detailed, multi-sentence user prompts and edit existing photos. It features a tool use capability: if a prompt lacks details, it can use a search tool to retrieve information from the web, and another built-in tool enables the model to generate code, such as a Python script to turn spreadsheet contents into a graph. Muse Image can loop in Muse Spark when its coding capabilities are insufficient. Meta says the model engages in “deliberate reasoning” before generating an image, enabling better use of infrastructure than the best-of-N method. Muse Image supports a test-time compute feature, and its self-refining behavior emerged during the reinforcement learning phase of training. On launch, Muse Image is available via the Meta AI chatbot in a limited number of markets and powers new image effects in Instagram stories. Meta plans to bring the model to Facebook, Messenger and more parts of Instagram in the future.