3D Animatic: Rough Previz in Your Browser, No Blender
An animatic is the rough render of a future video: object blocking, camera moves and timing. Now you can build one in the browser, with no Blender install, and hand the scene to Seedance.
Talking avatar from a photo: how to make an avatar video
Upload one photo, type a script or attach a voice recording, and get a talking video. A walkthrough of the Video avatars section on NeuralSpace: choosing an avatar, the engine, the frame format and the result.
Depth Map for Seedance 2.5: How to Get Past the Face Filter
Seedance 2.5 rejects recognizable faces and film frames. A depth map — black-and-white scene geometry — goes through. Here is how to make one free in your browser.
GPT Image API and MCP: how to add image generation to your
GPT Image 1.5, 2 and 2 Pro: capabilities, the /v1/images/generations endpoint, MCP tools and getting a key.
Open 9B model ZDTaichu5.0-9B wins 8 of 9 spatial benchmarks
A Chinese team open-sourced a 9B-parameter multimodal model. It leads every competitor under 10B on eight of nine spatial reasoning benchmarks and beats Gemini 3 Pro on MindCube-tiny.