DreamImage 2.0

DreamImage 2.0 is an image generation and natural-language editing model developed by Newport AI. Released in July 2026, the model is designed to facilitate instruction-based modifications of existing imagery. It bridges the gap between traditional text-to-image synthesis and precise manual editing by allowing users to describe desired changes in plain text while the system preserves the original image's structural composition and core content.

The model's primary capabilities include outpainting, inpainting, and colorization. It is also engineered for specialized vision tasks such as face swapping, background removal, and image enhancement. Unlike standard generative models that may deviate from source materials, DreamImage 2.0 is optimized to identify and modify specific elements or regions within a provided source image according to the user's textual instructions.

In industry benchmarking, DreamImage 2.0 has been recognized for its adherence to complex editing prompts. According to the Artificial Analysis Image Editing Leaderboard, the model achieved an Elo rating of 1,220 shortly after its release, positioning it as a significant tool for professional creative workflows. The model is primarily utilized within enterprise environments to streamline media production and design iteration through high-fidelity, prompt-driven image manipulation.