Open the model library

Choose Assets in the Overview sidebar and open the Models tab. The library belongs to the active team and is available on every plan, including Basic. Use search and the status or source filters to find uploaded and generated models.

Models count towards the team storage limit: the original or generated file, the Editor preview, reference images, and the thumbnail. A new upload or generation needs at least 200 MB of free team storage, and enough room for the model and its preview, otherwise the attempt is refused before anything starts. Delete unused models or other files to free space.

The summary shows model totals and processing state. Queued, Generating, and Processing are not ready for placement; Ready models can be inspected and used; Failed models need a corrected upload, prompt, provider, or retry.

Upload a model

Select New Model, choose Upload, enter a name and optional description, and select a self-contained textured GLB file. Review the file and submit it. MapToPlay keeps the original and prepares a lightweight Editor preview in a background task.

The upload is refused when free team storage is below 200 MB or cannot hold the model together with its Editor preview; the message names the space required and the space free. Once an upload is accepted, the storage it needs is held for it, so the preview cannot fail later because the team filled up meanwhile.

Open Tasks to follow processing. When the model becomes Ready, open it in the library, rotate the 3D preview, compare the original and optimized information, and use Download when you need the stored file.

Generate a model from a prompt

Select New Model, choose Generate, and pick an available model provider. If the provider requires a token, add or update it in the active team provider credentials and validate the available credits before continuing.

Enter the model name and a specific prompt between the shown minimum and maximum lengths. Add reference images up to the provider limit when they materially clarify shape, proportions, or style. Review the provider, prompt, references, and expected usage, then start generation.

Generation and preview optimization continue in the background. Watch Tasks and return to Models after the item is Ready. A generated result can still need revision for game scale, topology, materials, or visual style; inspect it before assigning it to many features.

Generation is refused before any provider credits are spent when free team storage is below 200 MB, because the finished model has to be stored. After it starts, the space it may need is held for it and the result is kept even if the team fills up in the meantime.

Generate from the Editor

In the Editor, select one polygon feature, open Integrations in the header, and choose the model-generation action. The selected footprint provides useful placement context, but the prompt and references still define what the provider creates.

The Access step needs a saved team token for the chosen provider. Select the token and use Validate token and credits once; Next unlocks when the balance covers the estimate. That result is remembered for the rest of the browser session, so reopening the wizard shows the earlier check and offers Check token and credits again when the balance may have changed. The token and the credits are checked again before the review step and before generation starts, so a remembered check never spends credits the team no longer has.

The result enters the same team model library and stays attached to the selected polygon while it is prepared. When it is Ready, the Editor places it at the model's original size. Use Select or Transform to correct supported position or rotation.

Manage and remove models

Each model costs 1 Task unit when its preview processing starts. When the team has no Task unit free, the model waits in the queue and starts as soon as units replenish; it is never rejected for that reason. While a model is still processing, the storage it reserved counts as used and is released when it becomes Ready or fails.

Open a model to review its source, status, description, original file, optimized preview, and failure message. Download before deleting when you need to keep a local copy.

Delete removes the library item for the team and can affect features that expect that model. Check Editor assignments and project use first. A failed or unwanted generation can be deleted safely when nothing references it.