At the recent "Force Original Power Conference," ByteDance's Volcano Engine unveiled two cutting-edge AI models: the DouBao Large Model 1.6 and the video generation model Seedance 1.0pro. This launch signifies a major leap in AI technology, particularly in multimodal understanding and video generation, showcasing formidable market competitiveness.
The DouBao 1.6 model has demonstrated exceptional performance in multiple international authoritative evaluations, especially in complex reasoning and competition-level mathematics tests. According to Tan Dai, President of Volcano Engine, this model not only supports multimodal understanding but also features a graphical user interface, enabling it to effectively tackle complex real-world tasks. Compared to its predecessor, DouBao 1.5, DouBao 1.6 has expanded its applications in e-commerce image recognition and autonomous driving data annotation. Its capabilities even extend to automating tasks like hotel reservations and organizing shopping receipts.
The release of the Seedance 1.0pro video generation model has also garnered industry attention. This model supports text and image inputs and can generate high-quality 1080P videos. It has achieved outstanding results in multiple evaluation rankings, surpassing competitors like Veo3 and Keling 2.0. As video content consumption continues to grow, this technology undoubtedly provides new impetus for the development of related industries.
Notably, the cost of using DouBao 1.6 has been significantly reduced, with pricing based on "input length." This allows enterprises to save substantial costs during usage. Specifically, for input ranges of 0-32K, DouBao 1.6's input cost is only 0.8 yuan per million tokens, with a comprehensive cost that is only one-third of DouBao 1.5. The cost of using Seedance 1.0pro is equally impressive, with the generation of a 5-second 1080P video costing only 3.67 yuan, making it highly competitive within the industry.
Furthermore, to support a broader range of AI applications, Volcano Engine has upgraded its AI cloud-native services. This includes the launch of MCP services, intelligent prompt tools like PromptPilot, and a series of new products, further promoting the implementation of intelligent applications. Hong Dingkun, Vice President of Technology at ByteDance, stated that with the continuous enhancement of AI capabilities, future development will become more efficient, and the barriers to entry will be lowered.