Reve just launched Reve 2.0, a new image generation model that ditches standard text prompts for structured layouts specifying each element's location, size, and attributes. Built on open-source Qwen models, Reve claims it trained with 10x fewer GPUs while hitting top arena spots among sub-$1T companies. Why it matters: separating layout planning from pixel rendering could give creators pinpoint compositional control.
Black Forest Labs just brought on legendary filmmaker Martin Scorsese as an advisor to guide its visual intelligence development with a focus on human craft and taste. The announcement came alongside a working storyboarding session where Scorsese tested its FLUX image generation model.
TripoAI just open-sourced TripoSplat, a lightweight tool that turns a single 2D photo into up to 260K 3D Gaussians under an MIT license. It runs on minimal code with ComfyUI support, though early tests show it works far better on basic objects than human faces.
Open-source frontend tool Impeccable just rolled out v3.5 to cure AI-generated design slop, compiling custom guardrails tailored to the specific design flaws of models like GPT-5.5 and Codex. It introduces a beta Live Mode for editing UI directly in the browser via voice or prompt and syncing straight back to source code, powered by a 20x faster CSS detector.
Researchers from Max Planck and CMU just open-sourced MAMMA, a markerless motion capture system that converts multi-view video into accurate 3D models with only four standard cameras. It tracks complex interactions like dancing and hugging without wearable sensors, easily pulling off Matrix-style bullet time effects. The code and a browser UI are available on GitHub for non-commercial research.
Nvidia just open-sourced SANA-Streaming, a 2B diffusion model that edits minute-long high-res videos in real time while they play. By combining linear recurrent blocks with attention, it hits 24 FPS at 720p on an RTX 5090 while keeping source motion locked in. Code and weights are now up on GitHub and Hugging Face.
Roblox just unveiled CubePart, a 3D generative model that outputs functional, pre-separated meshes tailored to a developer's defined schema. Instead of spitting out a single solid mesh, it generates accurately labeled parts like spinning wheels and opening doors that drop directly into physics and animation pipelines. It turns generative 3D from pretty digital statues into actual interactive assets.
Microsoft just released weights for Lens, a 3.8B text-to-image model that rivals larger 6B systems at a fraction of the training cost. Its distilled Turbo variant generates images in just 4 steps and 0.84 seconds on an H100, supporting resolutions up to 1440x1440. It brings high-end, lightning-fast generation directly to independent developers.
NVIDIA just introduced Gamma World, an interactive world model that scales generative simulation from single-agent setups to multiple players and robots. Operating at 24 FPS without pre-built geometry, it renders shared environments from coordinated multi-agent actions. Why it matters: multiplayer games and embodied robotics can now run entirely inside generative worlds.
CapCut just revamped its web creative suite with Design Studio 2.0, ditching rigid templates for an infinite AI canvas and a collaborative design agent. Creators can now circle-to-edit with brush tools, split layers automatically, and branch into new visual directions, backed by 20 free generations every day.
Runway just launched an MCP integration that lets users generate images and videos directly inside Claude, ChatGPT, Cursor, and Replit. Creators and developers can now tap into flagship models like Gen-4.5 and Kling without switching tabs or leaving their existing workflows. It essentially gives text and coding agents native access to top-tier creative media generation.
Leonardo.Ai's creative chief Dwayne Koh just tested OpenAI's GPT-Image 2.0 using plain, conversational ad briefs with no prompt engineering. The model delivered production-ready art direction and even spontaneously added a viral TikTok sticker to a Gen Z lip balm ad without being asked. It signals a shift where AI stops acting like a dumb rendering engine and starts behaving like a taste-aware creative partner.
Microsoft's MAI Superintelligence Team just launched MAI-Image-2.5, debuting straight into third place on the Arena leaderboard. The text-to-image model brings sharper text rendering, improved stylized illustrations, and better visual reasoning over lighting and space. It is live now on Arena, rolling out to MAI Playground and Azure Foundry in two weeks.
Nvidia just introduced LocateAnything, a visual grounding framework that lets you pinpoint and crop any object in an image using plain text descriptions. By decoding bounding boxes in parallel rather than token-by-token and training on a massive 138-million-sample dataset, it boosts dense object detection speeds by up to 10x while maintaining high localization accuracy.
Researchers just introduced TriSplat, a feed-forward 3D reconstruction model that builds scenes with oriented triangle primitives rather than Gaussian splats. By generating surfaces directly in a single pass in about 1.2 seconds, it exports simulation-ready meshes straight into physics engines and Unreal without post-processing headaches.
Researchers just introduced ControlLight, a low-light enhancement framework that lets you adjust illumination strength continuously without warping image structure. It uses a custom flow matching loss to keep visual details intact while brightening dark photos.
Researchers just introduced Pantheon360, a video diffusion framework that generates controllable 360° videos using an explicit 3D cache as a geometric scaffold. By decoupling 3D geometry from texture refinement, it solves the perspective drift common in narrow-view models, opening up cleaner pipelines for VR and digital twins.
Nvidia just introduced PiD, a pixel diffusion decoder designed to replace traditional VAEs and eliminate blurry high-resolution AI images. By turning decoding and upsampling into a four-step diffusion process, it renders 2K images from latents in under one second on an RTX 5090. Why it matters: It wipes out the slowest, most blur-prone bottleneck in modern image generation.
Researchers just unveiled GenRecon, a technique that reconstructs large-scale indoor 3D scenes from sparse RGB photos. By tiling overlapping chunks and lifting Trellis.2 generative priors to the scene level, it outputs editable PBR meshes and outperforms cutting-edge baselines by 16%. It brings casual smartphone snaps much closer to replacing expensive professional scanners for VR and game design.
Ideogram just launched an MCP integration that turns AI assistants like Claude and ChatGPT into design agents. Users can now generate images, build designs, and even train custom models without leaving the chat interface. It also works with Cursor and Hermes, embedding creative workflows right into your existing tools.
