Vidu Q4 Preview released, with references from 15 images and three audio clips
On October 7, Shengshu Technology released Vidu's next-generation model Q4 Preview on the web and through an API. The announcement time falls in the early hours of October 8 in Japan. It lists up to 15 image references, up to three audio references, and 10-bit output at 2K and 4K.
Coordination of expressions, voice, and movement and improvements to camera work are described by the company. The announcement lists introductory pricing from $0.014 per second, but notes that final prices and available resolutions and features vary by plan and region. The same unit price does not apply to every output.
Sources, images, and examples
SynthID Detector opens to the public, with checks for video watermarks
On October 7, Google announced the global public launch of SynthID Detector in English. It is a tool that uses image, video, and audio files to check for invisible watermarks added by AI from Google and its partners.
The official announcement names OpenAI, NVIDIA, and Kakao as partners, with Apple coming soon. Because it looks for supported watermarks, not finding one does not prove a file was made by a human.
Sources, images, and examples
SGF+ paper released, reporting continuous generation of up to 24 hours
The video generation study SGF+ was submitted to arXiv on October 7 at 17:09 UTC, in the early hours of October 8 in Japan. It proposes extending video by generating short segments in sequence, with separate training for the process that records past context and the process that generates the next visuals.
The authors report achieving up to 24 hours of continuous generation from training with five-second generations, without additional training on long videos. This is a research experiment, not a product specification guaranteeing consistent storytelling or characters across an entire 24-hour work.
Sources, images, and examples
Information as of October 8, 2026. Production ideas are the editorial team's analysis. Images and videos can be viewed in the linked original articles.
![October 12: AI video news roundup. [Reference] Runway Dev adds Grok Imagine Lite, with prices varying by resolution. Protoface API expands to more than 40 models, with multiple providers under one key. [Reference] Runway MCP uses Brand Kits assets as generation references. A character for the October 12 edition appears at the top right, and the 1 Minute Video Lab icon at the bottom right.](../../../assets/news/news-2026-10-12-mascot-12.webp)
