China’s ByteDance has released Vidi2, an AI video editor that can capture hours of footage and use it to generate a TikTok video or movie. Deedy Das, partner at Menlo Ventures, claimed that the video editor “understands video better than Gemini 3 Pro.”
Vidi2 is a multimodal large-scale language model with 12 billion parameters designed specifically for video understanding. It can reportedly process hours of raw footage, understand narrative structure, and generate complete TikTok short videos or movie clips based on simple prompts. This is considered a major disruption to the existing video editing industry.
This new model relies on fine-grained spatiotemporal localization (STG) capabilities that can simultaneously identify timestamps and bounding boxes of objects in videos. Given a text query, Vidi2 not only finds the corresponding time period, but also precisely marks the location of a particular object within that time period.
Read: US and China reach ‘final agreement’ to sell TikTok (
Byte Dance is developed Multiple practical automatic editing tools such as highlight extraction, story-aware cutting, content-aware layout restructuring, and multi-angle switching. All of these are based on Vid2 features and can run on consumer-grade hardware.
This technology has been applied to TikTok’s Smart Split feature, which automatically edits, restructures, adds subtitles, and converts long videos into short clips suitable for TikTok. Vidi2 can also turn simple prompts and trending topics into structured video titles, openings, and summaries.
According to AIbase, the release of Vidi2 and ByteDance’s TikTok (with 1 billion daily active users) benefits from a huge data platform provides massive amounts of video data for training and real-time feedback optimization, creating a huge challenge for native AI companies. Traditional AI companies may face greater competitive pressures as the technological flywheels of major platforms begin to turn.
Read: ByteDance eyes $330 billion valuation amid uncertainty surrounding TikTok in the US (
Vidi2 is still in the research phase and a demo will be released soon.
China-based ByteDance is best known as the parent company of the hugely popular social media platform TikTok. Recently, there were negotiations to sell TikTok’s U.S. operations from its parent company. US President Donald Trump is signed Last month, an executive order was issued to move forward with a deal for the TikTok app with a group of primarily American investors. The deal allows the app to remain online under bipartisan legislation passed in 2024 that would require it to divest its U.S. operations from its Chinese parent company.
Recently, ByteDance also launched AI assistant for mobile phones. This assistant will debut on ZTE’s Nubia M153 smartphone. The tool uses Doubao’s large-scale language model to handle voice tasks such as content search and ticket booking. ByteDance confirmed that it is in talks with multiple manufacturers to integrate the assistant into future smartphones. The company said it has no intention of developing its own hardware.
