Skip to main content
LocoPal AiLocoPal Ai
Quick Start

Guide

Platform Feature Guide

How to use the home page, model marketplace, API keys, logs, billing, asset library, and LocoPal Creation

View Markdown

Platform Feature Guide

This page introduces the main features of LocoPal Ai in the order you would use them day to day. After signing in, you only see the keys, tasks, assets, and bills of the current account.

Home page and model marketplace

The home page and the top product entries provide the console, user manual, model marketplace, and LocoPal Creation. The model marketplace presents model names, capabilities, parameters, and billing details in one place.

When choosing a model, confirm in order:

  1. The model's capability matches your business type, for example choosing a video model for a video task;
  2. The model ID appears in the /v1/models response of the current API key;
  3. The request parameters fall within the supported range shown on the model details;
  4. The estimated price fits your budget; the actual bill is authoritative.

Console overview

The console shows balance, plan, call volume, consumption trends, and recent records. The statistics help you identify high-consumption models and abnormal spikes, while the bill details let you verify the final cost.

API keys

An API key is the only credential for calling the model, image, video, audio, and asset APIs; the same key works across all APIs. We recommend creating separate keys for "Production Service", "Test Environment", and "Customer App", and setting clear names, quotas, and validity periods.

OperationDescription
CreateGenerates a new calling credential; once created it should only be stored on the server.
EditAdjusts the name, quota, model group, and validity period.
DisableTemporarily blocks this key from making requests; useful for stopping losses.
DeletePermanently removes a key that is no longer used.
Log filteringDistinguishes the call records of different applications by key.

Deleting or disabling a key does not delete the account's assets, tasks, or history; after creating a new key you can still use the original assets.

Usage logs and task logs

Usage logs record synchronous calls and can be filtered by time, API key, model, group, request identifier, and status. Task logs are for async business such as video and audio; after creating a task you should save the task ID. A change in waiting time does not mean re-billing; only creating a new task produces a new task identity.

Task logs show the submission time, task ID (including the upstream task number), elapsed time, status, progress, and details. After a task succeeds, a "Click to preview video" link appears in the "Details" column; click it to play the generated video directly in a pop-up, where you can copy the playback address without downloading it first. Administrators viewing other accounts' tasks in task logs can preview them the same way.

Logs only show the current account's own records.

Asset library management

The asset library is a collection of asset groups, offering two categories: virtual character assets and portrait assets.

Virtual characters

  1. Click "New Group", select "Virtual Portrait", and fill in the group name and description (both the name and description can be left empty; when left empty the platform names it automatically, for example "Virtual Portrait Asset 09-11 21:05");
  2. Within the group, click "Upload Image" to select a local image directly, or use "Submit Asset" to enter a publicly accessible http(s) image address;
  3. Asset submission is processed asynchronously; once the status shows "Available", the asset can be used reliably with models that support asset references;
  4. You can rename assets, refresh status, and delete them; before deleting a group you must first move or delete the assets in it.

When submitting an asset, the "Asset Name" can also be left empty; the platform names it after the uploaded file name. If you need a different name, just rename it in the asset list.

Portrait assets

Portrait assets must be captured and authorized by the person themselves on a mobile phone:

  1. Click "New Group" and select "Portrait Asset Library";
  2. Read and check the authorization agreement, then generate a QR code;
  3. Have the person themselves scan the code with a phone to complete capture, authorization, and profile entry;
  4. Return to the web page and click "Scan complete, start sync";
  5. The page shows the portrait group and the assets produced by the capture; if they are still being processed, refresh later.

The portrait group is bound to the current account that initiated the scan; other accounts cannot resolve or reuse that session. If you are told the feature is temporarily unavailable when creating an authorization session, keep the message and contact platform support.

Search and filtering

  • Search: both the group list and the asset list support searching by name or ID;
  • Filter: filter by group type and owning model to see only the groups and assets used by a particular model;
  • Owning models: the list shows only the first model name and a +N count; hovering over it lists all owning models of that group or asset;
  • Pagination: lists load in pages, with the total count at the bottom; you can choose 10 / 20 / 50 / 100 items per page (10 by default);
  • Refresh: re-reads the names, statuses, error messages, and access addresses of the current account's groups and assets.

Assets only need to be ingested once

You do not need to create a group or upload an asset separately for each model. On ingestion, the asset enters the platform's unified asset library, and every model that supports asset references uses the same copy directly; the asset ID stays the same, so you can reuse it across models without waiting for synchronization on the model side. For the correspondence between models and the asset library, see Asset Library API Overview.

Existing assets from before this change are prepared once by the platform the first time a model references them; after preparation they behave the same as directly referenced assets.

Per-library inspection (administrators)

At the top of the asset library management page, administrators can switch between the "Platform Asset Library" and the asset library of each model: after switching to a model's asset library, you can view the groups and assets in that library read-only (by default only the model name is shown), or create groups and submit assets directly in the library for special preparation and troubleshooting. Submitting assets in a library also supports uploading images directly; leaving the name empty names it automatically from the file name.

Balance, top-up, and billing

The balance page is used to top up and to view plans or redeem benefits. The model marketplace provides estimated prices, and the bill after a task completes provides the final settlement. When checking abnormal charges, use the task ID, request identifier, model ID, and time range; do not provide the full key.

LocoPal Creation

  • LocoPal Creation: a workspace for direct creation, offering entries such as Smart Mode, AI Drawing, AI Video, AI Audio, Short Drama, Works, Assets, and Prompts;

Creation reuses the same LocoPal account and model services; generated results, tasks, and assets are isolated per signed-in user.

Common workflows

API integration

Create an API key → query models → make a test call → verify in logs → integrate into your business service → set timeouts and error handling.

Multiple images and reference video

Prepare public media or assets whose status is "Available" → create an async video task → save the task ID → query until the terminal state → download or save the result elsewhere.

Portrait video

Create a portrait authorization session → the person scans the code to capture → resolve the session → query the portrait asset status → once the asset is available, submit a video task that supports the portrait asset library.

Was this page helpful?

On this page