跪拜 Guibai
← All articles
AI Programming

Tencent's WorkRally CLI Turns a Single Photo Into a Dubbed Short Drama for 30 Yuan

By 小虎AI生活 ·
Read original on juejin.cn ↗ Google Translate ↗ Alt translation

A CLI-driven AI video pipeline drops the cost of a four-scene dubbed short drama from roughly 8,000 yuan to 30 yuan and an afternoon. For solo creators and small shops, that turns video content from a budget line item into an impulse purchase, and it removes the production bottleneck that stops most people from starting.

Summary

WorkRally, built by Tencent's Hunyuan team, generates multi-view character sheets from a single uploaded photo and then keeps the character's face, clothing, and expressions consistent across all subsequent images and video clips. That solves the main tell that breaks AI-generated narrative content. The platform also exposes a command-line interface with 33 API commands, which means it can be driven entirely by an AI coding assistant like WorkBuddy.

A hands-on walkthrough shows the full pipeline: register, request 300 free trial credits from customer support, install the CLI, and feed the assistant a one-sentence prompt describing the story. The assistant then generates the character design, writes a four-act script, produces matching storyboard images, creates 5-second video clips per scene, and composites them with dubbing. The entire run took roughly three hours and consumed 300 credits, which cost 30 yuan at the platform's rate of 10 credits per yuan.

Traditional production for the same short drama would require an illustrator, a photographer, and an editor, costing around 8,000 yuan and coordination across three people. The cost comparison makes automated short-drama creation viable for solo operators who previously could not afford video production.

Takeaways
WorkRally generates multi-view character sheets from one photo and locks the character's appearance across all subsequent images and videos, solving the consistency problem that plagues AI-generated narrative content.
New users start with zero credits but can request 300 free trial credits from customer support; generating one image costs about 15 credits, and a 5-second video costs about 50 credits.
WorkRally provides a CLI with 33 API commands, making it scriptable by AI coding assistants for fully automated short-drama production.
Importing the WorkRally CLI documentation as a skill into WorkBuddy allowed a one-sentence prompt to trigger the entire pipeline: character design, four-act script, storyboard images, video clips, dubbing, and compositing.
A complete four-scene short drama consumed 300 credits (30 yuan) and roughly three hours of automated processing.
Traditional production of the same short drama would cost approximately 8,000 yuan and require coordinating an illustrator, photographer, and editor.
Conclusions

A CLI on an AI video platform is a force multiplier: it turns a visual creative tool into a programmable backend that an AI coding assistant can orchestrate end-to-end, which is far more scalable than clicking through a web UI.

The 300-credit trial is a deliberate on-ramp that lets users prove the pipeline works before they spend money, and the fact that credits come through a customer-service chat rather than an automated grant suggests the platform is still in an early, high-touch growth phase.

The cost delta between AI and traditional production is so large that the bottleneck shifts from budget to creative direction; the limiting factor becomes whether the operator can write a clear one-sentence prompt and judge the output, not whether they can afford to produce video.

Concepts & terms
WorkRally
A Tencent Hunyuan team platform for AI comic and short-drama creation that emphasizes character consistency across scenes and provides a CLI for programmatic control.
WorkBuddy
An AI coding assistant that can ingest documentation as a skill and then execute multi-step workflows by calling external APIs, in this case orchestrating WorkRally's CLI.
Character consistency
The ability to generate multiple images or video clips of the same character while keeping the face, clothing, and expressions identical, a persistent challenge in AI image and video generation.
Source: juejin.cn ↗ Google Translate ↗ Backup ↗