Tripo AI Releases Real-Time Generative Renderer EdenLive and Open-Sources It
Tripo AI has launched EdenLive, a real-time generative renderer dedicated to interactive scenes, as part of its interactive environment world model Project Eden, and it is now open-sourced to developers. EdenLive decouples scene evolution logic from visual rendering; rendering can be started with just a 2D/3D scene or reference video, a target first frame, and text instructions, and it supports both 480p and 720p resolutions; the 5-billion-parameter version can achieve real-time generation at over 24 frames per second at 720p on a single consumer-grade RTX 5090. The open-source release includes the 5-billion and 14-billion parameter model weights, real-time video generation code, the EdenGolem v0.1 skill for converting text into simple games, and a complete prompt-to-game front-end toolchain; the code is released on GitHub under the Apache 2.0 license, and the weights are available on Hugging Face.
Why read this
AI score 66
The original text specifies the open-source scope, parameter scale, and single-GPU real-time frame rate, allowing readers to assess the usability of real-time generative rendering in gaming and embodied AI scenarios.
Topics
Body · AI translation
English text is being prepared.
Tripo AI, which focuses on R&D of 3D-native foundation models and world models, has officially launched EdenLive, a real-time generative renderer dedicated to interactive scenes. The product belongs to its interactive environment world model Project Eden and is now fully open-sourced to developers.
The core design of EdenLive is to completely decouple scene evolution logic from visual rendering: code agents, game engines, or simulation engines govern the operating rules and changes of the scene, while the renderer converts raw low-precision frames into high-quality detailed video in real time, supports long-horizon continuous synthesis, and can continuously output real-time video content lasting several minutes.
To start rendering, users only need to provide a 2D/3D scene or reference video, a target first frame, and text instructions; it supports two resolutions, 480p and 720p. Measured data shows that the 5-billion-parameter version of EdenLive can achieve real-time generation at over 24 frames per second at 720p resolution using only a single consumer-grade RTX 5090 GPU. The product offers two operating modes: the real-time streaming mode outputs in chunked streaming form, suited to interactive scenes; the offline mode can process a complete reference sequence in one pass, better suited to re-rendering needs for existing videos and animations.
Its applications cover multiple fields. Paired with coding agents, it can be integrated into real-time preview workflows in game development; creators can quickly explore different visual styles based on the same scene layout, actions, and camera paths, with support for real-time replacement of characters and adjustment of overall visual tone; in the embodied AI field, it can convert real-person operation videos into robot first-person perspective footage, bridging the gap between human demonstrations and robot learning.
EdenLive is now officially open-sourced, with released content including the 5-billion and 14-billion parameter model weights, real-time video generation code, the EdenGolem v0.1 skill that enables text-to-simple-game generation, and a complete prompt-to-game frontend toolchain. The code is published on GitHub under the Apache 2.0 license, and the model weights are available on the Hugging Face platform.
Dr. Cao Yanpei, Chief Scientist of Tripo AI, stated that Project Eden splits world operation into a three-layer architecture: a continuously evolving world state, an interactive interface from state to visuals, and a generative renderer. EdenLive implements the latter two layers and does not restrict the functionality of the scene logic layer, making it an important step in the team's vision of building interactive generated worlds.
Source: auganix
Submissions/Tips: tougao@youxituoluo.com
Article/Business Cooperation: Lin Nan (WeChat 19250561593)
Join the industry exchange group: Lin Nan (WeChat 19250561593)
English text is being prepared.
Digital technology industry news and service platform
English text is being prepared.