Reve has released the beta version of its API, which allows for the modification of existing images using natural language instructions. The reve/edit-image model enables control over color, text, perspective, and other visual parameters using simple text prompts.
What Happened
The reve/edit-image model has been developed and is available through providers such as AIML API. The tool supports working with image URLs and allows for text queries up to 2560 characters long to make content edits.
Context
This release marks a shift from the "from scratch" image generation paradigm to a concept of controlled editing via natural language. The technology allows for the automation of complex visual tasks, such as changing lighting or removing objects, without the need for specialized graphic software.
Why It Matters for the Industry
For the industry, this means expanding content automation capabilities. Developers can integrate complex retouching functions directly into their applications, lowering the barrier to entry and enabling the creation of micro-SaaS services in design automation niches without deploying their own heavy models.
Why It Matters for Users
Regular users and content creators gain the ability to perform professional retouching and personalize visual content simply by describing the desired changes in words, without possessing skills in graphic editors.
What Is Not Yet Known / Limitations
At this stage, the product is in beta, so there is no precise data regarding latency, usage costs, or data security mechanisms.
Sources
Author
Look at AI, Editorial Team