AI video frontier shifts from 5-second clips to native 30-second takes
Independent-Date393 · reddit · 2026-08-11
AI video generation is undergoing a major leap: moving from the 5-8 second clip ceiling common just a month ago to natively supporting 30-second continuous takes.
- Tech Breakthrough: Seedance 2.5 shipped a native 30-second take, and Wan 3.0 public beta leads with the same capability. This means multi-shot scenes stop being a seam-hiding exercise and return to intentional cut lists, with notable progress in multi-person conversation continuity.
- Pricing Gap: Wan 3.0 beta is priced at $0.10/second for 720p ($3 for 30 seconds), whereas generating the same length via Seedance 2.5 API costs around $9.
- Input Expansion: Wan 3.0 directly reads structured files like doc, pdf, ppt, and even URLs, moving beyond traditional text prompts.
More from Multimodal
- First Fully AI-Generated Feature Film Open-Sources All 473,600 Assets for $2M — LinusEkenstam · 2026-08-11
- AI Video Fail: Recreating the Iconic Miami Vice Scene with Bert and Ernie — dreamwieber · 2026-08-11
- 10-Minute AI-Generated Film Goes Viral with 300K Views in a Day — ZabihullahAtal · 2026-08-11
- RefCaptioner: Precise Multi-Reference Image Grounding in Video Captioning — 机器之心 · 2026-08-11
- One Stylus Tap Turns Raw Ingredients Into Gourmet Ramen Using AI Video — SimplyAnnisa · 2026-08-11
- How to Identify AI Video Models by Their Default Duration — cocktailpeanut · 2026-08-11