At the FORCE Original Power Conference, Volcano Engine officially launched DouBao Large Model 1.6 and its new video generation model Seedance 1.0pro, marking a comprehensive upgrade of AI cloud-native full-stack services. As a leading family of multimodal, high-performance models, DouBao Large Model 1.6 has seen significant improvements in functionality and performance.
Breakthrough Pricing and Performance
In his speech at the conference, ByteDance's CEO Liang Ruobo announced that DouBao Large Model 1.6 adopts a unified pricing model, pioneering interval-based pricing based on "input length." Within the commonly used enterprise range of 0-32K inputs, the price is only 0.8 yuan per million tokens for input and 8 yuan per million tokens for output, reducing the overall cost by 63% compared to previous versions. Additionally, the pricing of Seedance 1.0pro is highly competitive, with each thousand tokens costing only 0.015 yuan, generating a 5-second 1080P video costing just 3.67 yuan.
#### Powerful Model Series
The DouBao Large Model 1.6 series consists of three models:
- doubao-seed-1.6: The first domestic model supporting 256K context, featuring deep thinking, multimodal understanding, and other capabilities.
- doubao-seed-1.6-thinking: Enhanced in deep thinking, supporting more complex reasoning capabilities.
- doubao-seed-1.6-flash: An ultra-fast version with extremely low latency, offering visual understanding capabilities approaching industry-leading levels.
These models not only significantly improve inference speed and accuracy but also support the implementation of more complex business scenarios.
Simultaneous Thinking and Search & Deep Research Functionality
The "simultaneous thinking and search" feature of DouBao Large Model 1.6 allows the model to conduct information retrieval while independently thinking. For example, when preparing daily necessities for a student from Guangdong living in Beijing, DouBao will deeply analyze the issue and recommend necessary items. The "Deep Research" function can generate detailed research reports in a short time, compressing days of work into 5-30 minutes.
Multimodal Understanding and GUI Operation Capability
DouBao Large Model 1.6 natively supports multimodal thinking, capable of understanding and processing complex real-world problems. This makes it highly practical in applications such as e-commerce product review and autonomous driving annotation. Additionally, the model boasts excellent GUI operation capabilities, enabling efficient interaction with various software for automatic task execution, such as hotel screening and invoice organization.
New Video Generation Model Seedance 1.0pro
The newly released Seedance 1.0pro video generation model features seamless multi-shot narrative, stable motion, and realistic aesthetics, widely applicable in industries such as e-commerce, film, and gaming. According to the latest authoritative third-party evaluations, Seedance 1.0pro leads in both text-to-video and image-to-video domains.
With these new functionalities and technologies, the release of DouBao Large Model 1.6 and Seedance 1.0pro marks an important step for enterprises in their AI intelligent transformation, laying a solid foundation for future innovative development.
