Tencent launches its answer to OpenAI’s Sora text-to-video tool

Chinese video gaming and social media giant Tencent Holdings has launched a text-to-video generation tool based on its Hunyuan artificial intelligence (AI) foundation model, joining an intensifying race among domestic rivals to challenge US peers in the field.

HunyuanVideo has been made available for free to enterprise and individual users, the company announced on Tuesday. With over 13 billion parameters – a measure of variables present in an AI system during training – HunyuanVideo is the world’s largest open-source model for video generation, the company said.

HunyuanVideo is designed to produce high-quality, “hyperrealistic” videos in which changing camera angles and reflections are rendered convincingly, Tencent said. In one demonstration video, a surfer rides a big wave and spins himself. In another clip, several hot air balloons rise slowly to the sky while campfires burn on the ground.

Tencent, owner of China’s most widely used social media app and the world’s largest video-gaming business by revenue, is the latest Chinese tech company to enter the AI battle in video generation, after ChatGPT maker OpenAI introduced Sora in February.

The Tencent headquarters in Shenzhen, Guangdong province. Photo: AFP

Short-video app operator Kuaishou Technology, the main rival of ByteDance in China, launched its text-to-video service Kling AI Model in June. E-commerce giant Alibaba Group Holding, owner of the South China Morning Post, is working on an AI video tool called Tora, which adopts OpenAI’s OpenSora as its foundation model.

Source link