AI Agent covers the full process of short drama production, and StarAI expands content production scenarios.
The StarAI project, an AI Agent creation tool, is further expanding the application scope of AI content production tools. Through continuous iteration of products and business, StarAI has gradually evolved from its early focus on AI short drama production to form an AI Agent workflow covering multiple core links of short drama production, and is extending to pan-commercial content scenarios such as AI e-commerce content.
StarAI is an AI content production company, whose core product is the self-developed AI Agent creation tool StarAI. It mainly serves AI short drama production teams, AI e-commerce content production teams, as well as B-end institutions and top creators with demands for large-scale content production.
Different from AI creation tools that only rely on a single generation capability, StarAI's product concept is to further embed the capabilities of large models into the actual content production process, and connect multiple production links that originally require manual work through Agents. The current product has independently developed core workflows including the one-click film production Agent, one-click style transfer Agent, and one-click short drama asset generation Agent, and is equipped with supporting production tools such as infinite canvas, covering multiple links from content creativity, visual assets to video generation and final film production.
According to StarAI's disclosure, more than 100 AI short drama production companies have accessed and used the service, covering about 1,000 corporate content producers. The average daily duration of AI content produced via StarAI is about 1,500 minutes. The project has achieved profitability, with monthly recurring revenue (MRR) stably maintained between 3 million yuan and 5 million yuan, and its annual recurring revenue (ARR) in 2026 is expected to exceed 30 million yuan.
From single-point generation to Agent workflow, AI short drama has entered the stage of large-scale production
With the rapid development of generative AI, the production mode of AI short drama is changing. In the past, creators mostly used a single AI model to complete image generation, video generation or copywriting creation. However, when content production enters the mass production stage, the generation capability of a single model can no longer cover the complete business process.
From early content creation to final film production, an AI short drama needs to go through multiple links including script decomposition, storyboard design, character and scene setting, character asset production, frame generation, image style transfer, video generation and post-production. Different links usually require the invocation of different models and tools. If the production process relies on manual switching between multiple platforms, it will not only increase the cost of learning and operation, but also easily lead to disconnection of materials, characters and visual styles across different links.
For AI short drama companies that need to continuously produce content, what really affects efficiency is not only how long it takes for a single model to generate an image or a video, but whether the entire production process can be standardized, automated, and further form a replicable workflow.
StarAI's product design is exactly based on this demand. Its core is not to retrain the basic model, but to combine the capabilities of different models with specific content production tasks, and complete task decomposition and execution in the production process through AI Agents.
At present, StarAI has independently developed the one-click film production Agent, one-click style transfer Agent and one-click short drama asset generation Agent. Among them, the one-click film production Agent is mainly used to improve the efficiency of the process from materials and scripts to the final video film; the one-click style transfer Agent meets the demands for visual asset style conversion; the one-click short drama asset generation Agent generates basic assets required for short drama production such as characters and scenes.
At the same time, StarAI also provides workflow tools such as infinite canvas, enabling different production links to be organized and collaborated in the same creation environment.
As a result, StarAI's product form is transforming from a simple "AI generation tool" to an "AI content production workflow". For short drama production teams, the value of the tool is no longer just providing an entry for invoking a certain model, but reducing switching between multiple tools and integrating originally scattered production tasks into a unified Agent workflow.
Model matrix + Agent workflow, covering a wider range of short drama production demands
In addition to the Agent workflow, model access capability is also an important part of the AI content production platform.
With the rapid iteration of video generation models, different models have differences in motion control, character performance, lens language, picture quality and generation speed. For professional content production teams, a single model is often difficult to meet all production tasks. Therefore, the ability to select and invoke different models according to specific scenarios has gradually become an important capability of AI content production tools.
StarAI is currently focusing on building multi-model access capability. According to the project disclosure, its platform has accessed the Seedance series of models, covering both domestic and overseas versions, including different model versions such as Seedance 2.0, 2.5, 2.0 Mini, 2.0 Fast, as well as the corresponding Seedance model services for the overseas market.
Public information shows that Seedance 2.0 has opened API services, supporting multi-modal inputs such as text, image, audio and video. It can be invoked through Volcano Ark in China, and the relevant model API services have also been provided through BytePlus in overseas markets. For StarAI, model access is not the ultimate goal of the product. More importantly, it further encapsulates different model capabilities into specific Agent workflows, so that users do not need to understand complex model parameters and invocation methods, but directly complete operations around specific production tasks.
This also makes StarAI different from simple model aggregation tools: the model is the underlying capability, and the Agent workflow is the middle layer connecting the model and the actual business. For AI short drama production teams, what users finally purchase is not a certain model itself, but a more stable and efficient content production capability.
Under this product concept, StarAI's current application scope has covered multiple links in AI short drama production, and is further extending to commercial scenarios such as AI e-commerce content. Scenarios such as e-commerce advertisements, commodity marketing videos, and brand content also have the characteristics of high frequency, mass production and standardized production, so they have certain commonalities with AI short drama production in the underlying workflow.
StarAI hopes to gradually expand its products to more pan-commercial content production scenarios through the Agent capabilities accumulated in the AI short drama scenario.
From more than 100 enterprise customers to 30 million yuan ARR, commercialization has entered the large-scale stage
In addition to product capabilities, StarAI's commercialization progress has also become an important change at the current stage of the project.
According to StarAI's disclosure, more than 100 AI short drama production companies have accessed and used the service, covering about 1,000 corporate content producers, and the average daily duration of AI content produced through StarAI tools is about 1,500 minutes. Compared with the early stage when it mainly focused on verifying product capabilities, StarAI has now formed a certain scale of enterprise customer base.
In terms of revenue structure, StarAI's current MRR mainly refers to monthly subscription revenue, which is stably maintained between 3 million yuan and 5 million yuan. According to the current business development plan, StarAI expects its 2026 ARR, that is, annual subscription revenue, to exceed 30 million yuan.
For AI content production tools, the continuous payment capability of enterprise customers is an important indicator to measure whether the product has truly entered the production process. Compared with the occasional content generation demands of individual users, enterprise customers usually have continuous video production tasks. Therefore, once the tool enters the daily production process, its subscription relationship and usage frequency are relatively more stable.
StarAI's current customers are mainly concentrated in AI short drama production companies, and it is also expanding to AI e-commerce content production teams, B-end institutions and top creators. The project hopes to further cover commercial customers with demands for large-scale content production through the product and technical capabilities established in the short drama scenario.
In this process, StarAI's business model is also gradually evolving from simply providing AI tools to a combined model of "subscription revenue + model invocation and related services". For enterprise customers, in addition to the Agent workflow itself, model invocation, content production efficiency and the adaptation capability between different models are also important factors affecting procurement decisions.
From the perspective of industry development trends, with the continuous improvement of AI video model capabilities, the focus of competition in the content production industry is gradually shifting from "who can generate videos" to "who can enable enterprises to produce videos continuously, stably and on a large scale". Under this change, Agent products that can further transform model capabilities into industry workflows have the opportunity to become an important tool layer connecting basic models and actual commercial scenarios.
The StarAI team is also continuing to iterate products around this direction. The core members of the team include Jiang Nan, co-founder in charge of product, commercialization and financing, Qi Xin in charge of AI short drama content production and commercial order expansion, and Ou Zhiyin in charge of AI Agent technology R&D.
Among them, Jiang Nan has previously worked at Sina, Wandoujia, Li Ning, Nenglian and other companies, with experience in Internet product, commercialization and team management. He has participated in the operation of products with tens of millions of daily active users and business with 100 million-level GMV, and has continuous entrepreneurial experience. The team has also previously participated in short drama content production and commercialization business, and has made practical exploration of the AI short drama production process.
From the perspective of product roadmap, StarAI does not limit its positioning to a single AI short drama tool, but hopes to take short drama as an entry point to continue expanding to a wider range of commercial content production scenarios. With the continuous improvement of Agent workflows such as one-click film production, style transfer and short drama asset generation, as well as the continuous access of multi-model capabilities, StarAI is trying to form an AI Agent product system covering more content production links.
For StarAI at the current stage, its core goal has further shifted from early product verification to large-scale commercialization: on the one hand, it continues to expand the coverage of enterprise customers, and on the other hand, it improves user usage depth through Agent workflows and model matrix, and further replicates the production capabilities that have been verified in the AI short drama scenario to pan-commercial content scenarios such as e-commerce and marketing.
If this model can continue to operate successfully, the competition of AI content production tools will no longer be limited to the generation effect of a single model, but will further enter the competition stage of workflow, scenario understanding and enterprise production efficiency.