AI tools for Rendering

SenseNova 6.8 Code2Video creates videos by generating executable production code. The released model focuses on short programmatic videos, including motion design and animated v

RoboSkills is an open collection of lightweight agent workflows and tools for robotics. It packages five working approaches: automated research, agent-as-policy control, agentic

Canva Craft is Canva Research's image-generation model for design-oriented imagery with deliberate layout and typography. A two-stage architecture plans element placement and te

InfiniSplat v2 reconstructs 3D Gaussian scenes from arbitrary collections of input views. Its dual-branch architecture freezes geometry while learning Gaussian attributes, using

Streaming Code World Model is a causal video-rendering implementation that turns a running proxy world into successive generated chunks. It uses a sliding six-chunk attention sp

StreamRender-H3 turns a code-controlled game world into a continuously generated visual experience. A code game engine maintains state and rules while a two-step streaming diffu

MOSS-Image is an image-model family with native layout control, typography generation, and a compact 5B backbone. It accepts ordinary text or JSON descriptions of objects, text,

Matrix-Game 4.0 combines executable game logic with streaming neural rendering for playable interactive worlds. A game engine owns state and rules, a coding agent handles new in

AgentGarten provides executable code worlds with a neural renderer for agents that learn through interaction. Programs own objects, rules, and goals, while a learned renderer co

OVOW, or One Video One World, reconstructs a monocular video as a structured 4D scene with separate mesh objects and motion. It targets simulation-ready output rather than only

LLaDA-Image is a six-billion-parameter model family for image generation and instruction-guided editing. A single checkpoint can create an image from text or alter a reference w

FixAnything repairs rendering artifacts from several kinds of 3D representation using a generalist video model. It refines degraded views from Gaussian splats, NeRFs, meshes, or

Code World Model uses a coding agent to update and control the state of a generated world. It separates persistent world structure from visual appearance, allowing the environme

Lucida reconstructs indoor captures as complete, editable object assets positioned in the observed scene. It separates the work into parsing, generation, and placement, aiming t

SenseNova-U1.5-8B-MoT is a native unified multimodal model for image understanding, text-to-image generation, and image editing. It is designed to improve visual quality, text r

FlaRe, or Floating Radiance Networks, is a research method for representing and rendering 3D scenes with a compact radiance field. The project focuses on floating radiance primi

Ideogram 4 is an open-weight image generation model built for design work, realistic 2K outputs, multilingual text rendering, and precise layout control. It is designed for crea

Qwen-Image 3.0 is the third-generation foundational image generation model in the Qwen-Image series. It focuses on rich content, authentic visual details, deep knowledge, and co

PhiZero is a research world model built around physical language, a compact discrete representation of world-state transitions learned from in-the-wild video. It separates physi

ViDiHand is a video diffusion model for reconstructing hand motion from egocentric and challenging video. It targets 4D hand recovery with robustness to occlusion, accurate pose