Meshy: 3D generator — what it does, formats, and rights to the result

Meshy is a cloud service that generates 3D models from text descriptions or images. The developer positions it as a tool for quickly obtaining ready-to-use models for games and 3D printing.

What it does

The service handles several tasks, which the authors describe as follows:

• Image to 3D — turns any image into a 3D model in seconds.
• Text to 3D — generates 3D models from text descriptions.
• AI Texturing — adds realistic PBR textures to any 3D model using AI.
• AI Image Generator — generates multi-view images optimized for 3D creation.
• Animation — automatically rigs and animates any 3D character in seconds.

The input is text or an image. According to the developer, the model is ready in under a minute. Export formats, as described by the author: GLB, FBX, OBJ, USDZ, STL. Download links are signed and time-limited. Polygons in the model:

• Remesh — from 100 to 300,000 (30,000 by default).
• Smart Topology — from 100 to 15,000 (4,000 by default).

Available model versions: meshy-6, meshy-5, meshy-t2, meshy-t1. For standard generation, meshy-5, meshy-6, and latest (corresponding to Meshy 6) are used. For Smart Topology generation — meshy-t2 (recommended by the developer, with cleaner topology, separated parts, and triangular output) and meshy-t1 (an older low-poly model that does not support setting the polygon count via target_polycount).

How to use it

The service works via an API. The developer describes it as a prepaid product: before use, you must purchase API access according to your application's needs. Additional API credits can be purchased on the subscription settings page. For enterprise pricing and custom contracts, the authors suggest contacting the sales department. The service does not provide its own model or model weights, and it does not run locally.

Rights to the output

Under the terms of use, the user owns the materials created through the platform, provided that no third-party copyrights were violated during generation. On paid plans, the rights to all created assets belong to the user. On the free plan, a CC BY 4.0 license is granted.

Who it suits

The service may be useful for those who need to quickly obtain a 3D model from text or an image without modeling skills — for example, for prototyping, populating game scenes, or preparing a model for 3D printing. Automatic texturing and animation capabilities expand the use cases. It will not suit those who need a local tool, access to the source code, or the ability to fine-tune the model on their own data. Polygon count limits and dependence on the cloud API may also be critical for tasks that require highly detailed geometry or stable operation without an internet connection. The developer has posted detailed documentation for each endpoint in the API reference, a link to which can be found on the documentation page.