HOOK12 Agnes 创作台

Author anonymousTotal Visits 21Visits in 7 Days 0

// preview

// description

Seven AI tools for vertical comedy, anime, video, text and images in one browser workbench

HOOK12 is a Chinese-language web workbench that gathers several AI generation utilities behind one dark-themed interface. The top navigation switches between seven pages. The 脱口秀 page builds a 12-second vertical stand-up clip from a single topic, with a selectable male or female performer shown as a photo card, a fixed 9:16 720P Mandarin lip-sync target, and a three-beat script structure. The 动漫 page produces an 18-second anime piece at 441 frames and 24fps from a premise, split into hook, development and ending. Three more pages cover general video: text-to-video with model, duration, aspect and resolution pickers; image-to-video with a list of reference image URLs; and a keyframe tool that takes first-frame and last-frame URLs plus a transition prompt. The final two pages do not need the video key: a text generator with an editable model field and optional system prompt, and an image generator with a model dropdown that pulls available models from the backend. Video pages ask for an Agnes API key stored in the local browser; text and image pages rely on a rotating server-side key.

Highlights

Seven AI tools

Video, text, image, anime

For vertical screens

Defaults to 9:16 clips

Explains the prompt

Shows each script plan

Kept in your browser

Own Agnes API key

Duration and ratio

Model and size selectors

Check by video id

Manual progress query

Chinese labels

Interface copy in Chinese

How to Use

Choose a tool tab
Open the site and use the seven links in the top bar to pick what you want to make: comedy, anime, text-to-video, image-to-video, keyframe video, text generation or image generation. Every page keeps the same header and swaps only the form below it.
Fill in the form
On the comedy page select the male or female performer card and type a topic; on the anime page describe characters, scene or plot; on the video pages choose a model, duration, aspect ratio and resolution and enter a prompt, and for image-to-video paste one reference image URL per line. The text and image pages also offer an optional system prompt.
Provide a key for video
The four video pages require an Agnes API key. Paste it into the key field at the top of the form; the page stores it in your own browser and only sends it when you submit. Without a key the page shows a reminder and the generate button cannot submit.
Generate and follow up
Press the generate button and wait for the result area or preview panel to update. The page fills a video id into the field below, and you can press the manual progress query button at any time to re-check that id instead of starting a new generation.