SenseTime Launches Short Video Platform Seko, Generating High-Quality Videos from Text

SenseTime’s newly-launched AI video agent Seko has attracted over 100,000 creators in Mainland China. Users can simply input text instructions and converse with the AI to generate videos, while maintaining consistent characters and styles across different scenes. By integrating creation and editing, this all-in-one platform streamlines the entire production process, reducing animation costs by up to 99.5% per minute.

Easing Creation Burden, and Accelerating Creativity

Seko is not a mere video generator; it is the industry’s first AI short video creation agent with integrated creation-editing capabilities, incorporating mainstream multimodal generative large models across the industry. It enables direct production of complete videos through dialogue.

Users don‘t need to worry about model selection, switching between platforms, or learning to operate different software. The process is so straightforward that users only need to communicate their creative ideas directly to output complete videos.

Pioneering "Creation-Editing Integration": Video Agent Outputs Complete Videos with One Click

While AI has greatly lowered the barrier to video creation, to produce a complete work still faces many challenges.

The tedious workflow involves "jumping back and forth" between multiple software tools—text-to-image, image-to-video, editing, and dubbing, each requiring a separate tool. This impacts efficiency and requires users to learn the rules of different platforms.

Learning prompts and adjusting parameters can also be frustrating. Modifying individual frames means starting over from scratch, resulting in repeated attempts and making the process uncontrollable.

SenseTime’s Seko is pioneering the industry’s "creation-editing integration" model, redefining video production workflows with a "true Agent." It provides a one-stop solution to the core challenge of how to efficiently create complete, coherent, and professional AI videos.

Any director knows that breaking down a script into shots is a significant challenge. Seko can accurately analyze complex scripts, deriving detailed shots and dialogue for different characters. On this basis, its powerful generation capabilities ensure that the visuals, lines, and dubbing for each character are accurately created and matched.

Furthermore, its innovative, professional-grade video editor offers an intuitive interaction experience, allowing users to efficiently edit and modify content based on a preview of the initially generated video before exporting their finished work.

Seko also supports shot replications and contextual memory for corresponding prompts. Combined with its image modification capabilities, it ensures the coherence of plots and scenes—all of which greatly streamline the creation process from text to video.

Solving the Consistency Issue, and Cutting Costs Drastically

The instability and inconsistency of AI content generation have long plagued creators. For time-consuming multi-shot videos, the characters, styles, and lighting often appear "random."

Following this month’s update, Seko has fundamentally solved this problem: it integrates the most powerful consistent image generation model, enabling the creation of ultra-long continuous shots with high consistency within a streamlined workflow. This capability elevates video production, powers the recent AI short drama trend, and establishes Seko as the 'AI Short Drama Master.'

Seko incorporates generative models such as SenseNova, Jimeng 4.0, Kling 2.1, Wan2.2, and Veo3. It intelligently matches the most suitable technology, allowing users to select any model without an additional workload for deployment.

Dubbing and background music are also fully covered: 60 natural-sounding voices are available for dubbing, and SenseTime’s self-developed SekoTalk technology supports unlimited-duration lip-syncing for long image-to-video conversions.

Seko not only ensures high-quality creative output but also drastically cuts costs. For example, in animation production, per-minute production costs can be reduced by 99.5% compared with traditional methods.

This smooth creation experience and cost efficiency have helped Seko gain over 100,000 users—including video teams, social media influencers, and popular short drama directors—with many citing it as the "best AI video agent tool."

AIGC creator Wang Zhijun used Seko to produce a short film themed around the “Flying to the Sky Project” (飛天計劃), which won the Best Creative Award at the AIGC Innovation Design Competition (古今雙飛天·AIGC創新設計大賽).