Skip to main content

Douyin's Duanbao Model Hits 30 Trillion Daily Tokens

Douyin's Duanbao Large Model Sees Unprecedented Growth

Beijing, October 2025 - Tan Dai, President of ByteDance's Volcanic Engine division, revealed staggering adoption metrics for the company's Doubao large language model during a recent industry event. The AI system now processes over 30 trillion tokens daily, marking one of the fastest enterprise AI adoptions in history.

Exponential Adoption Curve

The platform has seen token volume (the basic text units processed by AI models) skyrocket from:

  • 120 billion tokens/day (May 2024)
  • 3 trillion tokens/day (September 2025)

This represents a 253-fold increase in just 16 months, outpacing most competitors in the enterprise AI space.

"What we're witnessing isn't just technological adoption—it's a fundamental shift in how businesses operate," Tan stated during his keynote. "The Doubao model has become integral to workflows ranging from content creation to complex data analysis."

Driving Factors Behind the Surge

Industry analysts point to three key drivers:

  1. Vertical Integration: Seamless compatibility with ByteDance's ecosystem (including Douyin/TikTok)
  2. Cost Efficiency: Competitive pricing compared to Western alternatives
  3. Localization: Superior Chinese language processing capabilities

The model now serves applications across:

  • Automated content generation
  • Intelligent customer service systems
  • Real-time data analytics platforms
  • Marketing campaign optimization

Strategic Implications

Volcanic Engine has positioned Doubao as an "AI-as-a-Service" solution, allowing enterprises of all sizes to integrate advanced capabilities without massive infrastructure investments. This approach appears particularly effective in China's rapidly digitizing SME sector.

The company reports that over 60% of users employ the model for multiple simultaneous use cases, suggesting deep integration into business processes rather than experimental deployments.

Future Roadmap

Tan hinted at upcoming multimodal enhancements that would combine:

  • Text processing
  • Image generation
  • Video analysis

expected to launch Q1 2026.

Key Points:

  • Daily token volume grew from 120B to 3T (+253x) in 16 months
  • Serving diverse industries including e-commerce, finance, and media
  • ByteDance leveraging its ecosystem advantage against competitors
  • Multimodal expansion planned for early next year

Enjoyed this article?

Subscribe to our newsletter for the latest AI news, product reviews, and project recommendations delivered to your inbox weekly.

Weekly digestFree foreverUnsubscribe anytime

Related Articles

News

ByteDance Doubles Down on AI Hardware with Douyin Phone 2 and Smart Glasses

ByteDance is making waves in the AI hardware space with its upcoming Douyin Phone 2, set to launch mid-2026. The tech giant is taking a collaborative approach, partnering with smartphone manufacturers while expanding into wearables. Their ambitious plans include two versions of AI glasses - one lightweight model arriving soon and a more interactive version later this year.

January 29, 2026
ByteDanceAI HardwareSmartphones
News

TikTok Doubles Down on Shenzhen with New AI and Video Tech Hub

ByteDance's TikTok is expanding its footprint in China's tech hub Shenzhen with a second headquarters focused on AI and video technology. The Nanshan District facility will house research labs and business incubators, complementing TikTok's existing Greater Bay Area operations. This move signals the company's growing investment in southern China's innovation ecosystem.

January 8, 2026
ByteDanceShenzhenTechAIInnovation
News

ByteDance's DouBao AI Glasses Set for Limited Release

ByteDance is gearing up to ship its highly anticipated DouBao AI glasses, but with a twist - the first batch of 100,000 units will be exclusively available to existing DouBao App users. Powered by Qualcomm's Snapdragon AR1 chip, these lightweight glasses focus on audio functionality without a display screen. While the company remains tight-lipped about broader sales plans, industry insiders reveal development is already underway for a second-generation model.

January 6, 2026
wearable techartificial intelligenceByteDance
News

ByteDance Shuts Down Rumors About Doubao AI Glasses Launch

ByteDance has officially denied rumors about the imminent launch of Doubao AI glasses, stating there's no concrete sales plan yet. Despite speculation fueled by supply chain reports, the company maintains the product hasn't reached shipping phase. Industry experts suggest this reflects ByteDance's cautious approach to hardware launches, possibly due to technical challenges or strategic timing considerations.

January 6, 2026
ByteDanceAI WearablesSmart Glasses
News

ByteDance's Doubao AI Glasses Set to Shake Up Wearable Tech Market

ByteDance is making waves in wearable tech with its upcoming Doubao AI glasses, developed in partnership with Longqi Technology. These sleek glasses feature a custom chip for seamless voice interaction and translation, while keeping weight and price low. Unlike standalone devices, they smartly leverage your smartphone's power through the Doubao app. Priced under 2000 yuan, they could make AI glasses mainstream accessories rather than niche gadgets.

January 5, 2026
wearable technologyAI hardwareByteDance
ByteDance's StoryMem Gives AI Videos a Memory Boost
News

ByteDance's StoryMem Gives AI Videos a Memory Boost

ByteDance and Nanyang Technological University researchers have developed StoryMem, an innovative system tackling persistent issues in AI video generation. By mimicking human memory mechanisms, it maintains character consistency across scenes - a challenge even for models like Sora and Kling. The solution cleverly stores key frames as references while keeping computational costs manageable. Early tests show significant improvements in visual continuity and user preference scores.

January 4, 2026
AI video generationByteDancecomputer vision