豆包率先抢跑付费赛道 留给千问、元宝的定价空间不多了

Doubao, the AI-native app owned by ByteDance with the largest monthly active user (MAU) base in China, has officially launched a full paid subscription for its professional version. Below is a summary of key takeaways:

1. **Feature Differences**: Powered by the top-tier Doubao 2.1 Pro large language model, Doubao Professional reaches production-level performance in three core areas: programming, AI agents, and vision-language models. Its core upgrade is a dedicated office task mode that can independently decompose work tasks. It supports local device operations, has built-in Office tools, multi-domain skill invocation, scheduled tasks, and small application generation, enabling users to directly deliver finished work outputs and meet professional productivity demands.

2. **Pricing & Options**: The professional version adopts tiered pricing, with the entry standard tier priced at 68 RMB per month — far lower than ChatGPT Plus’s 135 RMB monthly fee. Meanwhile, the original free version remains fully functional and will continue to be updated. Casual everyday users can still use it for free, while heavy professional users can choose to upgrade to paid services based on their needs.

Doubao is an early mover in China’s AI paid services track, providing a reference case for commercial operation, pricing strategy and user management for domestic AI brands. Key takeaways are summarized below:

1. **User Segmentation and Consumption Trends**: AI users in China are now clearly segmented: ordinary users only need to meet basic daily question-and-answer needs, while heavy users for professional productivity scenarios consume far more tokens than general users and are willing to pay for advanced capabilities. User分层运营 is an inevitable trend for AI brand development.

2. **Pricing and Brand Strategy**: Doubao’s tiered pricing, set far lower than overseas competitors, allows it to pre-empt domestic users’ mental budget for AI paid services. By keeping the free version updated continuously, it protects the experience for ordinary users while separating high-consumption users to control computing costs, resulting in a very clear overall strategy.

3. **Risk Warning**: After launching paid services, Doubao’s MAU saw its first decline. Beta test users also feedback that the gap between the paid and free versions is not large enough, and personalized features are lacking. This shows brands need to refine user experience in advance and clearly differentiate product tiers before launching paid services.

Doubao’s early entry into the AI paid track has sent a clear market signal to domestic AI service sellers. Below is a summary of opportunities and risks:

1. **Market Opportunities**: China’s AI paid market has officially launched, and users have clear willingness to pay for AI productivity tools. The segmented commercialization model has already proven its preliminary viability: separating high-consumption professional users from free general users helps control high computing costs while meeting the needs of users at different tiers, making the model feasible.

2. **Risk Warning**: Doubao’s 68 RMB per month entry-level price point has already squeezed pricing space for future competitors, significantly compressing the room for new entrants. In addition, if the experience gap between paid and free versions is too small, it will trigger user dissatisfaction, leading to low conversion rates, user churn, and even damage to user confidence in the entire category.

3. **Best Practices to Learn From**: Sellers do not need to rush to pursue paid penetration. In the early stage, they should focus on verifying the functional value of the product, advance commercialization gradually, and reduce transformation risks.

Doubao’s launch of a mature paid AI productivity tool offers relevant insights for factories pursuing digital transformation and efficiency improvement opportunities. Key takeaways are summarized below:

1. **Available Business Opportunities**: AI tools have now crossed the threshold to production-level usability. Doubao Professional’s capabilities in data organization, data analysis, scheduled tasks, and process automation can be directly applied to factory operation and management, helping factories cut costs and improve efficiency. Factories can introduce such tools to quickly advance digital upgrades for office and operation work.

2. **Insights for Digital Transformation**: AI can now independently decompose tasks and invoke multiple tools to complete work. Factories do not need to invest heavily in proprietary in-house development to access mature advanced AI capabilities. They can use ready-made AI tools to quickly implement digital transformation and lower the barrier to adoption.

3. For factories planning to build their own AI services, they can reference Doubao’s cost control approach: segment users and scenarios by usage frequency to balance costs and user experience.

Doubao’s launch of its paid professional version reveals new development trends in China’s AI service industry. Key takeaways are summarized below:

1. **Industry Development Trends**: China’s AI industry has shifted from the early stage of fighting for free traffic to the commercialization stage. Product positioning is also shifting from general-purpose question-and-answer tools to productivity tools, and professional AI services that can directly deliver work outputs will be the core growth driver going forward.

2. **Core Pain Points for Clients**: On one hand, AI vendors face enormous pressure from computing costs: daily computing costs for leading products can reach tens of millions of RMB, making a pure free model unsustainable. On the other hand, the core pain point for C-end paid users is the insufficient differentiation between paid and free versions, a lack of personalized features, underwhelming experience, and difficulty improving conversion rates.

3. **Viable Solutions**: Players can adopt a segmented user operation model: separate heavy high-consumption professional users from ordinary daily users via a paywall, launch advanced models and productivity features targeted at professional users, adopt tiered pricing to lower the entry barrier, and prioritize verifying functional value in the early stage rather than rushing to hit monetization targets.

As a leading AI-native platform launching a full paid system, Doubao provides many references for AI platform operation and commercialization. Key takeaways are summarized below:

1. **Changing User Demand**: AI platform users are no longer satisfied with basic free Q&A, and have clear advanced demand for professional productivity scenarios. Platforms need to launch corresponding products and services to meet these demands and seize the opportunity for commercial transformation.

2. **Updated Best Practices to Reference**: Adopt a segmented user operation strategy: keep the free version updated continuously to meet ordinary users’ needs, and launch a paid professional version to serve heavy high-consumption users. This approach controls high computing costs while covering the needs of users at different tiers. For pricing, adopt tiered pricing lower than overseas competitors to pre-empt users’ mental budget for AI paid services and build a first-mover competitive advantage.

3. **Risk Mitigation Points**: Conduct sufficient beta testing before launching paid services, and expand the experience gap between paid and free versions to avoid user dissatisfaction and churn. Do not take paid penetration as the core performance indicator in the early stage; verify functional value first before advancing commercialization to reduce transformation risk.

Doubao’s early launch of a complete paid subscription system is an important milestone for the commercialization of China’s large model industry. Below is a summary of key insights for research:

1. **New Industry Trends**: China’s leading large model products have officially exited the phase of competing for user traffic and entered the stage of commercialization and implementation. The industry’s focus has shifted from growing user scale to deepening productivity value, and AI products are transforming from traffic products to productivity tools, with subscription-based paid models becoming the mainstream commercialization direction.

2. **New Industry Challenges**: Infrastructure investment for the large model industry is extremely high: ByteDance plans to invest $70 billion in AI infrastructure by 2026, and leading products incur tens of millions of RMB in daily computing costs. The original free model can no longer cover these high costs, while user acceptance of paid services remains uncertain. Issues including declining MAU and underwhelming beta experience have already emerged, so the commercialization path still needs further validation.

3. **Reference for Business Model Innovation**: Tiered subscription uses a paywall to separate users with different needs, which controls computing costs while protecting the experience of users at all tiers. It is an innovative exploration for the large model industry to cope with high cost pressure, and offers important reference value for business model design across the entire industry.