SuperAPI AI Directory

LibTV hands-on test – one-click script storyboard generation, integrated with mainstream models such as Seedance 2.0

Author: AI小集 · Original source: AI工具集

I woke up and my lobster had learned to make short dramas!

Recently a lot of friends have been asking if there are any projects that can make money using lobsters. I really did figure out a path: using lobsters combined with LibTV to make short dramas.

Just now,LiblibAI launched a brand-new content creation platform LibTV, letting lobsters also call models on their own and orchestrate workflows to generate short dramas!

No need to understand storyboards, no need to know how to score music, and no need to design characters. In minutes, your lobster can transform into a professional director.

 

01. Hands-on test of LibTV

 

LibTV has 2 ways to use it. We can directly create workflows manually on the canvas, or generate workflows through the lobster, to complete the production of text, images, videos, audio, and scripts.

Official website:https://www.liblib.tv

LibTV's operation interface is an infinite canvas, connecting each piece of content through nodes.

LibTV has four preset workflow templates to help us quickly get started creating content.

We only need to master 4 main functions to master LibTV:

1. Double-click on an empty space to add a node. There are five types of nodes: text, image, video, audio, and script. Just choose the one you need and create it.

2. Click an image, and the toolbar above will have many image processing tools, such as HD, multi-angle, nine-grid, etc.

3. We can box-select multiple nodes, so that multiple images can be used together as references to generate what we want.

4. For the workflows we have connected, after modifying elements, we can run them again in batches.

Let's try making a short film ~ choose story script generation.

Enter the script:

"I Write the World in the High Tang"

Genre: ancient style / time travel / power-fantasy comic drama

Suggested duration: 60–90 seconds

Tone: passionate × epic feel of the High Tang × satisfying rhythm

[Prologue]

[Modern · Late-Night Office]

The sound of keyboard typing is urgent. The computer screen's blue light is harsh. Shen Zhaozhao (female, 28 years old), pale-faced, bends over the desk working overtime. Phone notification sounds keep coming. She looks up, her vision blurs— Black screen.

[Act One]

[High Tang · Golden Throne Hall · Day]

Drumbeats shake the heavens. Shen Zhaozhao suddenly opens her eyes, already kneeling in the center of the hall, dressed in male official robes, her hands restrained. Officials stand in ranks, the palace towering. A minister says sternly: "Treason! Falsely claiming poetic talent that shocks the world, deceiving the emperor, you should be beheaded!" Shen Zhaozhao says in shock, softly: "This is... the Tang Dynasty?" Above the dragon throne, the emperor's gaze is deep. Emperor: "Since you claim unparalleled talent, compose a poem in the hall. If you cannot—behead." The gleam of blades reflects in her pupils. The great hall is deathly silent. [Act Two] Shen Zhaozhao slowly rises. Her breathing gradually steadies. She looks outside the hall— Chang'an is bathed in sunlight for ten thousand li. She says softly: "Since I have come to the High Tang... then I will contend with the High Tang." She raises her head and recites loudly: "Do you not see the waters of the Yellow River coming from heaven—" Her voice echoes through the hall. The assembled ministers are shaken. The emperor slowly stands. The sound of the poem is impassioned, like thunder falling upon the hall. [Act Three] Outside the hall, wind rises. The streets and markets of Chang'an are lined with endless lights. The emperor says softly: "This poem shall be passed down for ten thousand generations." Shen Zhaozhao's gaze is firm. She murmurs to herself: "Since I have come... then I will write all thirty thousand li of Chang'an." The camera pulls away. The mountains and rivers of the High Tang unfold.

LibTV will automatically help us break down the storyboard script.

The duration of each shot, what kind of image it shows, which characters are included, and what shot type is used are all recorded crystal clear.

We add a node to batch-generate storyboard images. On LibTV, we can use almost all mainstream image generation and video models. The latest Seedance 2.0 is also available now:

The character consistency in the storyboard images is maintained quite well. If we're not satisfied with a particular storyboard, we can click to select it and regenerate it.

We can batch-generate storyboard videos from the storyboard images.

For generated storyboard images and storyboard videos, we can modify the prompt and regenerate them. Clicking on a storyboard image will pop up a prompt box for us to modify.

When modifying a storyboard image, if you only want to use a local part of the image for regeneration, click the image and select the multi-grid split in the top right corner,

After splitting, you can select the desired block to use as a reference image to generate a new storyboard image.

The multi-angle feature that just launched a few days ago can now also be used when editing images~ We can regenerate multi-angle views of an image without adjusting the spatial layout.

Video after modifying the details:

The visuals are done very well~ However, the transitions between scenes need a bit more fine-tuning, which would make it more suitable for a continuous series.

On LibTV, there are also tons of creators who have published their own workflows, and we can reuse them with one click.

For example, for this promotional short film production, we only need to replace the product images with our own, and then rerun the nodes:

A brand-new promotional short film is ready. The quality is very good.

 

02. Connect LibTV to Lobster

 

First, you need a Lobster~ It can be local or in the cloud. Here I'm using AstronClaw, and deployment is also very simple. For details, you can check out the hands-on tutorial I posted.

First send Lobster a command to install LibTV API Skills. Lobster will automatically install this Skill.

Help me install LibTV API Skills, address:https://clawhub.ai/haofanwang/libtv-skill

Click the top right corner in LibTV to create an Access key.

Send the Access key to Lobster, and then Lobster can generate all kinds of videos.

Give me a 30-second comic drama telling a worker's version of "Covering One's Ears While Stealing a Bell"

Lobster will keep an eye on the progress for us at any time.

Let's take a look at the final generated result:

03. Some Sharing

 

LibTV is absolutely an epoch-making product. It integrates the most comprehensive video and image models on the internet, and also comes with a large number of exclusive creation features. From script writing and storyboard drawing to video generation, editing, and image optimization, the complete content creation process can all be done on the same canvas.

LibTV covers almost all the steps needed to make a comic drama, and each step also has a lot of practical small features.

Whether it's multi-camera 9-grid scheduling, character three-view settings, or visual timeline deduction, these extremely hardcore features are truly productivity tools that hit the pain points of film and television production, advertising planning, and short drama developers.

LibTV packages uncontrollable AI compute into a professional film and television industry pipeline. AI creation is transitioning from single-point blind box draws to a fully controllable process. The opening of APIs and workflows means that future video creation may be a mode of collaborative operation between humans and Agents.

Original article link:Outrageous! My Lobster learned to make short dramas? LibTV is simply a comic drama神器