ByteDance Builds an AI Video Tool Matrix: Jimeng, Xiaoyunque, Comic Drama Creation Tool, and Jianying Hub

By XinSheng Pro, Author | Feng Yong (Shanghai), Producer | Zhang Yitong (Shanghai)

In mid-August, ByteDance beta-tested an Agent for the comic drama industry, the "Comic Drama Creation Tool"; on September 21, the overseas brand BytePlus launched "Dramagic," a short drama production Agent for overseas markets; during the Douyin Creator Conference, Jianying released Jianying Hub, an AI video creation entry point that connects AI generation and editing.

Looking at products at the application layer, ByteDance's ecosystem has now built an AI video tool matrix including Jimeng, Xiaoyunque, the Comic Drama Creation Tool, Dramagic, and Jianying Hub, achieving systematic segmentation for different content domains and production scenarios.

Jimeng AI, launched in May 2024, is the earliest AI video generation platform introduced under ByteDance. Currently, Jimeng AI's core functions are generation and canvas. Generation covers images, video, audio, music, digital humans, and motion imitation, among other areas, while the canvas serves as a director's workbench for professional creators, offering an extremely high degree of manual adjustment and enabling precise control over details such as character costumes, spatial positioning, and camera movement.

It is evident that what Jimeng pursues and showcases is the upper limit that AI technology can achieve, benchmarking against "film-grade creation." Within Jimeng AI, the skills displayed toward the front are also a series of functions essential to premium AI content creation, such as cinematic long-take camera movement, storyboard creation, and micro-expression control.

Since its launch, Jimeng AI has also consistently been ByteDance's core platform for connecting with top-tier premium creators. Starting in 2025, through a series of initiatives such as the Future Image Plan and the Young Directors Collaboration Plan, Jimeng AI has discovered a group of AI super-creators and brought AI super-creator works to international film festivals including Cannes, the Shanghai International Film Festival, and Busan.

Since the beginning of this year, Jimeng's incubation and support for high-quality creators and works has upgraded from technology and computing power to the construction of a more systematic, comprehensive ecosystem.

The most representative example is Jimeng Studio, a film and television content label launched in August this year. Continuing Jimeng AI's long-standing support for film-grade creation, Jimeng Studio is soliciting works for film and series projects. Beyond technology and computing power, in the distribution and commercialization stages, Jimeng Studio will also assist selected projects in connecting with theatrical circuits, offering revenue mechanisms such as box office revenue sharing, joint distribution, or guaranteed minimum revenue sharing, while also supporting brand sponsorship, IP derivative development, and cross-media cooperation, and assisting with submissions to major domestic and international film festivals such as Cannes, Berlin, and FIRST, and planning offline premieres, themed screenings, and customized marketing plans. Jimeng Studio's first batch of partners includes traditional leading companies in the animation and film and television fields, such as Shanghai Animation Film Studio, YHKT Entertainment, Chen Liaoyu Studio, Huanyu Film, and Wanzhong Hexing Film and Television.

In February, the launch of Seedance2.0 improved the quality of AI video generation while further lowering the creative threshold for ordinary users. Jimeng AI's mobile app, Jimeng APP, began to show stronger standalone product attributes, building a UGC AI content community aimed at the general public—in simple terms, Jimeng APP is attempting to create an AI version of Douyin.

Through the AI avatar feature, ordinary users with no AI creation experience can also obtain a creative video of their own. During the Spring Festival, Jimeng AI partnered with "La La Land," launching an AI creative challenge for its tenth-anniversary re-release and releasing officially authorized materials. At the same time, it also invited celebrities such as Li Dan, Li Xiaolu, and Huang Yi to join the community.

Within the Jimeng APP interface, in addition to the "Following" and "Recommended" tabs, what is rather distinctive is the "Inspiration" tab, where users can directly extract other creators' prompts and content to make the same style of video, which is very convenient for ordinary users to get started. Currently, when opening Jimeng APP, the mainstream content is also highly fun and entertaining AI UGC: AI costume changes, AI travel photography, AI song and dance, or becoming a domineering CEO charging through a hail of bullets, or cultivating immortality and flying on a sword across the four seas.

The recruitment details of the Jimeng AI Ad Remake Contest held in July can demonstrate the directional differences between the Web and mobile ends. The Web contest targets hand-crafted creative works, recruiting premium creative blockbusters >12 seconds, with plot twists, creative imagination, and cool visual effects; the APP contest targets performance and remake works, <12 seconds in length, using avatar performances to generate creative short videos, with one-tap follow-along shooting, brainwashing performances, and creative secondary edits.

Also incubated by the Jianying team, the product turning point of Xiaoyunque is likewise closely tied to the launch of Seedance2.0. In February, because users could experience the latest Seedance2.0 model, Xiaoyunque's user volume surged.

In March, Xiaoyunque launched a short drama Agent, the industry's first short drama Agent equipped with the Seedance2.0 model. At that time, AI short dramas were at a full-scale breakout moment, and Xiaoyunque attracted industry-wide attention with its practical batch delivery capability: without needing reference characters, storyboard sheets, or editing, uploading a script and story setting would automatically complete parsing and batch generation, supporting one-click film completion for scripts of up to 100,000 characters.

Over the past period, Xiaoyunque has also conducted much technical exploration for emerging film and television products such as "live-action + AI" and interactive works. In June, Xiaoyunque collaborated with Huanyu Film to co-produce "The Beauty," exploring "live-action shooting + AI," in which costumes, makeup, props, environments, and the integration of real people with the environment were all produced by Xiaoyunque; in August, Xiaoyunque and the Seedance 2.0 interactive model were applied to the interactive short drama "Global Freeze: I Built a Doomsday Safe House."

Beyond the short drama Agent, another core feature of Xiaoyunque is the marketing Agent. The Xiaoyunque marketing Agent supports uploading character and product reference images, selecting suitable creative concepts, hooks, and styles, and then generating content; viral creative styles and Douyin-side data are connected.

In both the short drama and marketing directions, Xiaoyunque's customers are mostly organizations with batch content production needs, and the core features of the Agent are designed to simplify workflows and improve overall efficiency.

Xiaoyunque's positioning is undergoing a new change. On August 31, Xiaoyunque AI announced a brand upgrade, with "Create good stories together" becoming its new positioning; a relevant person in charge of Xiaoyunque AI said: "Xiaoyunque AI hopes not only to help creators generate a clip, but to accompany creators in completing a story with characters, plot, and a worldview." This change did not happen overnight; in fact, since April, Xiaoyunque has already launched multiple rounds of incentive programs for high-quality story-type OPCs.

English text is being prepared.

On the other hand, in August, ByteDance internally tested a more vertical short drama Agent, the "Comic Drama Creation Tool," which is said to be powered by multiple underlying models from Doubao, with the slogan "Make creation purer."

The Agent is a tool built by ByteDance for the comic drama track, targeting creators and content companies, especially for batch and project-level comic drama generation. The interface is very simple; upon opening the homepage, there is only a script upload function; after uploading a script, through a one-click selection of four steps—project planning, script outline, asset design, and episode planning—the episodes can ultimately be generated. The originally built-in options for different large models have also been replaced by clearer generation requirements more targeted at specific plots.

The biggest innovation of the Comic Drama Creation Tool is shown in its mass production capability: scripts support up to 200,000 characters, and team collaboration supports up to 30 people. At the same time, its ecosystem is interconnected and bound with ByteDance's short drama ecosystem; upon logging into the system, creators are prompted to bind a short drama creator center account.

A veteran comic drama practitioner shared her user experience with "Sanshidai": "The Comic Drama Creation Tool is mainly aimed at Hongguo short drama producers; it supports one-click film completion, can batch-produce character assets, and after parsing the script it even writes the prompts for you, and can directly run 100 episodes at a time." She also revealed that in Q3 of this year, a friend in the industry used the Comic Drama Creation Tool to achieve monthly revenue sharing of several million yuan.

Another AI video tool being advanced in parallel is Dramagic, aimed at the overseas comic drama market, launched on September 21 this year by BytePlus, a brand under ByteDance's overseas business.

Its positioning and approach are similar to the Comic Drama Creation Tool: through four stages—script, assets, storyboard, and video generation—it can go "from script to preview," emphasizing its industrial assembly-line capability and making it more suitable for team and enterprise use; it is essentially ByteDance's Agent for respectively laying out the domestic and overseas comic drama markets.

In addition, at the Douyin Creator Conference in September, ByteDance also brought AI capabilities into its editing tool CapCut, launching multiple new AI features such as "CapCut Hub" and "CapCut Assistant."

CapCut Hub interface

CapCut is strongly bound to the Douyin ecosystem, and as an editing tool it has very high penetration and usage among the general public, making it one of the core tools for connecting and empowering individual creators to produce UGC content on Douyin.

The biggest breakthrough in this product update is connecting AI generation and AI editing. CapCut Hub combines an infinite canvas and an Agent, building in the generation and editing workflows for text, images, audio, and video, with a process similar to creating in an Agent: splitting the script, planning storyboards, saving assets, and generating previews. The convenient part is that by clicking "More editing," you can directly enter CapCut's usual multi-track editing timeline without importing and exporting again.

CapCut Assistant's efficiency-boosting role is even more prominent, as it can largely replace manual rough cuts and automatically handle issues such as slips of the tongue, repeated sentences, and overly long blank spaces. After the rough cut is completed, the smart packaging feature provides video templates that can be applied directly to the editing timeline. In addition, there are many built-in features such as cutting spoken narration, commentary, and AI text-to-video.

In related discussion posts on Xiaohongshu, some netizens commented that "credits don't burn fast" and "even I, a novice, already find it very powerful." For most users, "AI + CapCut" can already solve video generation and editing challenges to a large extent in one stop, and is expected to attract more users to participate in UGC content creation in the future, broadening the community's content pool.

Overall, in the AI video track, ByteDance's product layout is very dense. AI video content as a whole is developing toward premium quality, and AIGC creators are also flowing among different content themes. Objectively speaking, there is inevitably overlap in features and ecosystems among ByteDance's various AI video tools at present, and there has recently been news that some businesses of Jimeng and Xiaoyunque are about to merge.

But we believe that ByteDance's current layout in the AI video tool field is not entirely the result of internal horse racing; at the foundation, ByteDance's content ecosystem is large enough and differentiated enough. Different platforms such as Douyin, Hongguo, and PineDrama involve different content such as short videos, short dramas, medium- and long-form videos, and overseas short dramas, covering various types of creative teams from OPCs and studios to film and television companies. Completely different production scenarios and content needs impose requirements on AI tools at different dimensions.

Using different products to meet creative needs in different creation scenarios lays a foundation for ByteDance's currently growing AI content in multiple fields. This itself also conforms to the development pattern of AI tools, evolving from a single general-purpose tool further toward vertical Agents for different production scenarios and different service targets.

It is worth noting that although what ordinary users can most feel at present is the tool upgrades in entertainment content fields such as film and television and animation, the continuous improvement of large model capabilities is further penetrating more physical industries. At the end of July, Seedance 2.5 was officially released, and on the day of release, multiple companies including XCMG, XPeng Motors, Lingchu Intelligence, Weifen Zhifei, and Qiongche Intelligence confirmed cooperation and will take the lead in introducing the model's capabilities.

Currently, these companies have already begun exploring scenario implementation based on the Seedance series models. For example, XCMG applies it to industrial operation training and process SOP video generation, while Qiongche Intelligence uses video generation to expand training data for embodied intelligence large models, generating operation data suitable for different machine types without relying on physical robots to collect one by one, thereby improving the model's adaptability and generality across multiple robots. Extending from the content industry to scenarios such as manufacturing, retail, robotics, and autonomous driving, video generation models are beginning to become standardized production tools in business processes.

Wang Shuai, associate professor in the Department of Computer Science and Engineering at the Hong Kong University of Science and Technology, believes that video generation models are moving from generating content to understanding the world, and from assisting creativity to enhancing productivity.

This means that the development of technology not only points to the reshaping of content production methods and supply sources, but is also becoming a new technological foundation for the development of more physical industries.