Guide
LocoPal Creation (Creation Platform)
Create images, videos, audio, short dramas, assets, and works directly with LocoPal Creation
LocoPal Creation (Creation Platform)
Entry: https://creation.locopalai.com
Creation is a workspace for creators: it breaks the process from "idea → finished piece" into several reversible steps and reuses LocoPal Ai's models, billing, and assets. It does not require creating an API key; when you are not signed in, it completes authorization through the LocoPal unified sign-in page.
Main features
| Module | Purpose |
|---|---|
| Creation Agent | Enter a sentence (a creative idea, script, or visual requirement), and the Agent breaks it down and schedules models to generate |
| Text, image, and audio/video modes | On the home page you can directly choose Smart Mode, AI Drawing, AI Video, or AI Audio |
| Canvas | Put multiple generations on one canvas for repeated iteration and recomposition |
| Short drama | Organize shots and assets by episode; suitable for continuous storytelling |
| Works / Assets | Archiving of generation results; assets interoperate with the LocoPal Ai asset library |
| Prompt library / Lexicon | Accumulate reusable prompts and styles |
Sign-in
Creation has no separate account. When you open creation.locopalai.com without being signed in, it automatically redirects to locopalai.com/sign-in; after signing in you return to Creation without entering the password again.
Choosing models
The "Select Model" panel in the creation interface is grouped by capability:
- Text: understands requirements, and writes scripts and prompts;
- Image: generates and edits images;
- Video: text-to-video and image-to-video;
- Audio: text-to-speech, narration, and dubbing; the specific capabilities depend on the model catalog.
Checking a model enables it. The source of these models is exactly the same as LocoPal Ai's model catalog — whatever is enabled in the backend groups can be selected here; after adjustments, refresh the page for them to take effect.
Billing
All generation uses the LocoPal Ai account quota, and prices are determined by group pricing. View top-ups, plans, orders, and consumption records in the wallet of the LocoPal Ai console, and verify each call in "Usage Logs / Task Logs". Creation itself does not require creating an API key.
Recommended workflow
- State your requirements clearly: in the input box, describe the subject, scene, purpose, and the desired aspect ratio or duration;
- Choose a mode: when unsure about parameters, use "Smart Mode", and the Agent will break down and plan;
- Add references: when you need to keep the person, product, or style consistent, upload a reference image or select an existing asset before submitting;
- Submit and wait: view the task status on the current page, and do not resubmit a task that is still running;
- Save the result: download it when finished, or reference the result in the next round of creation.
Prompts should ideally cover the subject, action, shot, visual style, ratio, duration, and explicit constraints. For video tasks, first confirm the resolutions, durations, and reference asset modes supported by the model; unsupported combinations return a parameter hint when the task is created.
How to choose reference assets
| Requirement | What to do |
|---|---|
| One-time use this time | Upload directly in the input box; used only for the current submission |
| Long-term reuse | First ingest into "Asset Library Management", wait until the status is "Available", then reference |
| Need a real person's likeness | The person must first complete QR code authorization; an ordinary upload cannot replace it |
Temporarily uploaded assets are not automatically ingested. For content you want to use again in other conversations, projects, or videos, first ingest it and wait until the status becomes available; see Asset Library and Portrait Assets for details.
Generation status
| Status | Description |
|---|---|
| Queuing | The task has been accepted and is waiting to run |
| Generating | The model is being called; you can leave the page and return to the original record later |
| Completed | The result has been saved; download it or reference it in later creation |
| Failed | The original error information is kept; fix the assets or parameters and retry |
Refreshing the page does not create duplicate tasks; reopening the page restores the most recent records and task statuses from the server.
FAQ
Does it say "model unavailable" when generating a video? First check the model marketplace of LocoPal Ai to confirm whether the model has been enabled under a group; if you just adjusted the groups, refresh the page and check again.
Insufficient balance? Top up on the wallet page of LocoPal Ai. Creation does not charge separately.
Do I need to create API keys? No. Creating with Creation directly does not require a key; you only need one when integrating the capabilities into your own system, see Integrate the API and Create a Key.
Will refreshing the page regenerate? No. Refreshing only restores the page state; it does not recreate tasks or bill again.
It says "failed to read reference asset"? This is usually not a prompt problem, but the model service cannot read the reference asset. First confirm on the asset card that the status is "Available", then see Video Generation and Retry.