豆包养不起2亿用户

This article outlines the current development and commercial status of Doubao, ByteDance's AI chatbot product:

1. Doubao has already become a mainstream consumer AI application in China, with daily active users (DAU) ranging from 140 million to 200 million across different data tracking metrics. However, unlike traditional internet products, every AI conversation consumes computing power. Doubao incurs tens of millions of yuan in daily computing costs, while its daily revenue is less than 1 million yuan, leading to annual losses in the tens of billions of yuan from computing costs alone.

2. To curb its losses, Doubao is set to launch a paid membership tier with monthly prices ranging from 68 yuan to 500 yuan, and the highest annual membership costs 5088 yuan. It is also being integrated with Douyin E-commerce and local lifestyle services to drive traffic to Douyin Mall. While ByteDance has denied plans to reduce functionality for free users, Doubao still lost 6.1 million monthly active users in May 2026.

3. Doubao frequently faces issues with incorrect AI outputs, such as wrong ticket refund fee information and failed generation of correct steamed bun images. This has led to widespread online discussion about its "Doubao personality"—a polite AI that repeatedly makes mistakes.

This article analyzes the current state of AI product commercialization and provides key insights for brands looking to build AI-enabled businesses and tap into consumer trends:

1. Consumers are extremely sensitive to AI product pricing. Even when developers explicitly promise not to cut free features, launching paid tiers can still trigger user churn. Brands developing AI-related products should design pricing and commercialization strategies carefully to reduce user resistance.

2. ByteDance is integrating Doubao with e-commerce and local lifestyle scenarios, and Doubao will emerge as a new public domain traffic entry point. Brands can prepare to connect with this new channel early to capture new customer acquisition and conversion opportunities.

3. To-C AI monetization is still in the exploratory phase globally, while B2B services and AI video generation have already proven profitable. Brands can proactively explore these areas, develop AI-powered marketing and design tools aligned with their core businesses, and gain an early foothold in the space.

This article discloses ByteDance's commercial layout for its AI business, and provides clear opportunity alerts and risk warnings for e-commerce sellers:

1. On the opportunity side, ByteDance is integrating Doubao with Douyin E-commerce and local lifestyle group buying scenarios, and Doubao will become a new traffic entry point for Douyin Mall that supports direct in-app checkout. Both e-commerce sellers and local lifestyle merchants can monitor Doubao's merchant onboarding policies and prepare early to capture this new traffic source.

2. On the risk side, consumer acceptance of paid AI remains low, and AI is inherently prone to factual errors. When using Doubao for marketing or customer service, sellers must conduct thorough information checks in advance to avoid reputational and financial damage caused by incorrect AI outputs.

3. AI monetization business models are still immature, so sellers should not blindly invest massive resources into AI initiatives. It is sufficient to monitor developments and test the waters gradually.

This article analyzes the current monetization landscape of the AI large model industry, and offers insights for factories advancing digital transformation and deploying AI applications:

1. The operational logic of AI applications is fundamentally different from that of traditional internet products: larger user volumes translate directly to higher computing costs, and even a product with 200 million DAU can face massive losses. When advancing digitalization and rolling out AI applications, factories should not blindly chase traffic. Instead, they must calculate cost projections clearly first to avoid unsustainable overinvestment.

2. The market has already validated that B2B enterprise services and AI video generation are clear, profitable directions. ByteDance's Seedance AI video generation model posts an annualized revenue of around $2 billion with a 70% gross margin. Factories with relevant technical capabilities can expand into B2B AI services to unlock new business opportunities.

3. Cost-effective AI models better match market demand. ByteDance's newly launched Seedance 2.0 mini cuts costs by 50% and expands profit margins. When deploying AI applications, factories can also start from a cost-reduction perspective, prioritizing use cases with clear costs and confirmed monetization paths.

This article sorts out the core pain points and trends of China's AI large model industry, and provides guidance for AI service providers to identify promising market directions:

1. The core pain point facing the industry today is the difficulty of monetizing consumer-facing large models, where traditional traffic-based monetization formulas no longer work. The entire industry is still exploring new business models, which has created massive market opportunities for B2B AI service providers.

2. The market has already confirmed that B2B enterprise services and AI video generation are the clearest profitable paths in the current AI industry. ByteDance's Seedance AI video generation model already generates over 1 billion yuan in monthly revenue with a 70% gross margin. AI service providers can strategically focus on these two areas to meet existing market demand.

3. The top pain point for customers today is the high cost of using AI. ByteDance's low-cost, high-performance Seedance 2.0 mini model has been well received by the market. Service providers can develop targeted cost-reduction solutions to meet the needs of small and medium-sized customers and expand their customer base.

This article discloses ByteDance's commercial exploration of its AI business, and offers key insights for the operation and development of AI platforms:

1. On the operational side, the free-traffic business model is no longer sustainable for AI platforms, which must explore reasonable pricing models. Doubao's paid membership trial is a pioneering move for the entire industry. However, platforms must note that users are highly sensitive to AI pricing, so it is critical to communicate clearly with users before launching paid plans and address concerns about free-tier functionality cuts to avoid large-scale user churn.

2. On the monetization side, platforms can explore integrating AI with their existing core scenarios. Doubao's integration with Douyin E-commerce and local lifestyle services to drive e-commerce traffic is a model worth replicating for platforms with existing scenario resources, as it allows them to fully unlock the value of their existing traffic.

3. On the risk side, AI is prone to generation errors that can trigger public relations crises. Platforms need to improve error accountability and model optimization mechanisms to reduce reputational risk. They can also actively expand merchant recruitment for B2B services and AI video generation to diversify their revenue streams.

This article provides an in-depth analysis of Doubao's C-end monetization challenges, showcasing the latest developments in China's large model industry and offering high reference value for industrial research:

1. A new paradigm shift is underway in the industry. The traditional internet rule "traffic equals revenue" no longer holds in the AI era: every AI inference call consumes GPU computing power, so larger traffic leads directly to higher costs. Even Doubao, a leading AI application with 200 million DAU, posts annual losses of over 10 billion yuan, which is a shared structural challenge for the entire Chinese AI industry.

2. The industry's exploratory direction is gradually clearing up: while C-end monetization remains highly uncertain, the industry is increasingly turning to B2B enterprise services and video generation models for viable paths. ByteDance's Seedance AI video generation model has achieved stable profitability that can offset Doubao's losses, validating the feasibility of this direction.

3. Doubao's trial of paid membership represents an industry-wide exploratory step for AI pricing models. The resulting user reaction provides a real-world case study for research on user acceptance of AI products and pricing strategy, and the industry as a whole will gradually develop new business models adapted to the AI era.