Geo News
Community curated by people like you
LatestAICryptoHealthWorld AffairsUS Politics
ByteDance and MiniMax release competing AI video models within hours
00

ByteDance and MiniMax release competing AI video models within hours

Jul 31, 2026

Chinese AI firms ByteDance and MiniMax released competing video generation models, Seedance 2.5 and H3, within hours of each other. ByteDance's model can generate 30-second clips with audio, while MiniMax is challenging closed-source rivals by releasing H3's model weights. The dueling releases highlight China's growing leadership over the US in the burgeoning AI video arena.

MiniMax H3 model launch

  • ▪MiniMax's H3 model can generate 4-15 second video clips at 2K resolution with native stereo audio
  • ▪MiniMax launched H3, a general-purpose multimodal generation model that processes text, images, video, and audio as a unified context
  • ▪MiniMax's H3 is available via an application programming interface (API) and its consumer-facing Hailuo AI platform

Performance benchmarking results

  • ▪H3 trails Google's Gemini Omni Flash in text-to-video tasks and ranks behind ByteDance's Seedance 2.0 and Gemini Omni Flash in image-to-video capabilities
  • ▪According to benchmark platform Artificial Analysis, MiniMax's H3 is the world's most powerful AI model for video editing

Open-source weights strategy

  • ▪The H3 weights will be released under the MiniMax Community License, allowing free non-commercial use and commercial use for firms with under US$20 million in annual revenue
  • ▪MiniMax plans to release H3's full model weights to challenge the "dominance" of closed-source models in video generation

ByteDance Seedance 2.5 release

  • ▪Users can provide Seedance 2.5 with up to 30 images, 10 video clips, and 10 audio files as references for generation
  • ▪Seedance 2.5 is available on the Jimeng AI and Doubao Pro platforms, with API access planned through BytePlus ModelArk
  • ▪Seedance 2.5 can produce video clips up to 30 seconds long, which is longer than Google's Gemini Omni Flash's approximate 10-second limit
  • ▪ByteDance released Seedance 2.5, an AI video model that generates video and audio in a single pass
  • ▪MiniMax Group Inc. and ByteDance Ltd. released updates to their AI video generation models within hours of each other

Commercial applications

  • ▪MiniMax's H3 is designed for commercial content creation in film, advertising, video games, and product design
  • ▪Director Neill Blomkamp used the previous version, Seedance 2.0, to create a 13-minute short film called "Nightborne."
  • ▪ByteDance showcased Seedance 2.5's capabilities in a short film titled "The Missing Pair," which was produced entirely with the model
  • ▪The releases from the two Chinese firms are seen as underscoring the country's leadership over the US in the AI video generation field

4 sources

Marktechpost
MiniMax Releases MiniMax H3: An Omni-Modal Video Model That Generates 15-Second 2K Clips With Native Stereo Audio
View source article
Bloomberg
China’s MiniMax and ByteDance Release Dueling AI Video Models
View source article
Scmp
Video AI: MiniMax challenges ByteDance with low price, open weights for new H3 model
View source article
The-decoder
ByteDance's Seedance 2.5 generates 30-second video clips with built-in audio
View source article

Featured stories

View more in AI video generators

OpenAI and Synopsys partner to develop AI model for chip design

Sep 30, 2026 · 3 sources

DeepSeek releases software tools for Huawei AI chips to challenge Nvidia

Sep 30, 2026 · 4 sources

OpenAI launches Dots, always-on AI agents that work across 4,000+ apps

Sep 29, 2026 · 14 sources

OpenAI unveils ChatGPT overhaul with shared workspaces, plugin system, and $500 monthly tier at DevDay

Sep 29, 2026 · 13 sources

Story comments

Loading comments…

Related entities

MiniMaxChina

Related Projects

ByteDance

Topics

AI video generatorsText to videoAI tools & productsChinaGenerative audioVideo generation modelsSeedance 2.0

Featured stories

View more in AI video generators

OpenAI and Synopsys partner to develop AI model for chip design

Sep 30, 2026 · 3 sources

DeepSeek releases software tools for Huawei AI chips to challenge Nvidia

Sep 30, 2026 · 4 sources

OpenAI launches Dots, always-on AI agents that work across 4,000+ apps

Sep 29, 2026 · 14 sources

OpenAI unveils ChatGPT overhaul with shared workspaces, plugin system, and $500 monthly tier at DevDay

Sep 29, 2026 · 13 sources