Help & Reference
This page unifies the terms that appear in the documentation and the platform interface, so the same concept is not called different things on different pages.
| Term | Meaning |
|---|
| LocoPal Ai | The API platform and the core of the entire ecosystem: it provides accounts, sign-in, billing, the model catalog, the asset library, and task records. Entry: locopalai.com. |
| LocoPal Creation | The creation platform (web workspace), aimed at direct creation without writing code. Entry: creation.locopalai.com. |
| Console | The management interface after signing in to LocoPal Ai, including the model marketplace, API keys, logs, asset library management, and wallet. |
| Term | Meaning |
|---|
| API keys | Credentials for calling the platform APIs, in the form sk-.... One key can call the model, image, video, audio, and asset APIs. |
| Access Token | A short-lived access token. The platform frontend keeps it only in memory and does not write it to localStorage. |
| Refresh token / session cookie | Used to restore the signed-in state after a page refresh. Its scope is .locopalai.com, so both platforms can share sign-in. |
| Unified sign-in | Sign in only once to LocoPal Ai, and the other platforms automatically receive the signed-in state. LocoPal Ai is the single source of identity. |
| Signed-in state invalidation | After signing out, switching accounts, or changing the password, sessions on both platforms are invalidated at the same time, and you need to sign in again. |
| Term | Meaning |
|---|
| Model ID | The value to put in the request's model field, from data[].id returned by GET /v1/models. It is not the name shown on the page. |
| Model marketplace | The entry point for viewing model names, capabilities, parameters, prices, and available groups. |
| Call group / group | The authorization scope that determines which models a key can see and call. A group contains several channels, and each channel carries several models. |
| Channel | The path under a group that actually carries models. Whether a model is available depends on whether it can currently be routed to an available channel. |
| Base URL | The API prefix, such as https://locopalai.com. Write only the domain; do not append /v1 again. |
us-api.locopalai.com | An alternate access address for overseas models or specific network environments. Whichever address you use, its /v1/models is authoritative. |
| Synchronous call | A call where the result is obtained as soon as the request returns, such as text, image, and audio. |
| Async task | A call where you first create a task to get an ID, then poll for results, such as video generation. |
| Term | Meaning |
|---|
| Asset library | The platform's unified asset storage, organized by group. Once ingested, an asset can be reused by every model that supports asset references. |
| Asset group | A container for assets, divided into two types: virtual (virtual) and portrait (real). |
| Virtual asset | An uploaded or generated image asset, suitable for saving reusable materials such as characters, products, and styles. |
| Portrait asset | A real-person asset that must be liveness-captured and authorized by the person themselves using a phone to scan a QR code; it cannot be replaced by an ordinary image upload. |
| Asset ID | The unique identifier of an asset, consistent across models. |
asset://<asset_id> | The syntax for referencing an ingested asset in a request. Use it only after the asset status becomes Active. |
Active | One of the asset statuses, meaning the asset has finished processing and can be reliably used with models that support asset references. |
Processing | The asset is still being processed. This is not an error; continue polling with the same asset ID. |
data_url | The form for submitting image data inline (data:image/...;base64,...). Only one of it and image_url can be submitted. |
| Term | Meaning |
|---|
| Task ID | The identifier of an async task. It must be saved after creation; even if the local wait times out, track it by querying rather than recreating. |
| Task logs | The entry point for viewing the execution details of async tasks, including submission time, task ID, elapsed time, status, progress, and artifacts. |
| Usage logs | The entry point for viewing the details of synchronous calls, including model, tokens, and cost. |
queued / queuing | The task has been accepted and is waiting to run. |
processing / generating | The task is running; you can leave the page and return to the original record later. |
succeeded / completed | The result has been generated and can be downloaded or referenced. |
failed / failed | The task has failed with a terminal status. Read error.code and error.message in the same response. |
video_generation_failed | The terminal failure code for video tasks; error.message is already a cleaned-up, readable reason. |
| Artifact | A task output: an image, video, or audio file that can be previewed or downloaded directly. |
| Request snapshot | The task's request content and original request JSON. Visible only to administrators; this field is not returned when ordinary users query their own tasks. |
| Term | Meaning |
|---|
| Token-based billing | Input and output tokens are billed separately; the unit is shown as 1K or 1M tokens. |
| Cached input / cache write | Billing items shown only for models that support prompt caching. |
| Per-call billing | Each successful request is billed at a fixed price. |
| Price matrix | The price table that video models show by "whether video input is included" × "output resolution". |
| Settlement on completion | The actual charge for an async task occurs when the task completes; submission only accepts the task and does not pre-charge. |
| Wallet | The entry point for top-ups, plans, orders, and billing records. |
| Estimated consumption | The cost estimate before a task is submitted. If the balance is insufficient to cover it, a message is returned directly and no task is created. |
| Term | Meaning |
|---|
Rate limiting / 429 | Requests are too frequent. Back off and retry as instructed; do not retry concurrently. |
task_concurrency_exceeded | The concurrent task limit has been reached. Keep the existing task IDs, wait, and let the user actively resubmit. |
| Backoff retry | After a failure, wait a period before retrying, extending the wait each time, instead of resending immediately and repeatedly. |
| RequestId / request ID | The identifier of a single request, used to locate the specific record when troubleshooting. |
400 parameter validation failure | The request was not accepted and usually does not incur a generation task charge. |
