MeshifAI

What is MeshifAI?

MeshifAI is a text-to-3D platform designed for developers and 3D content creators; users describe objects in natural language, and the system generates downloadable GLB models. It is suitable for use in games, WebGL, AR/VR, educational purposes, and rapid prototyping, but it is not a general-purpose code generation tool.

The product offers a JavaScript SDK as well as a Unity workflow; generation can be triggered from within the application code, or it is possible to enter instructions in the Unity editor, preview the results, and save the assets. This document focuses on the creation of simple models, while for more professional results, it may still be necessary to carry out cleanup and optimization tasks in Blender, Unity, or Unreal.

Main functions

• Text-based 3D generation: Enter an English description of the object, character, or environment, and the service will create a corresponding 3D mesh. The description should include information on the main elements, shape, style, proportions, and key details, avoiding any conflicting requirements.
• Texture-free model: The default path generation is faster and more stable, making it suitable for verifying outlines, volumes, and scene placeholders. Materials, UVs, textures, and animations can be added later using professional tools.
• Texture models: It is possible to request models with PBR materials, which offer a higher level of visual quality; however, the documentation indicates that these models are highly experimental, require more time to generate, and have lower reliability.
• Control of creative variation: The SDK offers options such as variance to adjust the degree of difference between various generated results. A higher level of variation is suitable for exploration, while when precision is required, the variation should be reduced and reproducible inputs used.
• GLB output: The results generated currently are in the GL Transmission Format Binary format, and can be used in Blender, Unity, Unreal, web viewers, as well as in most AR/VR workflows.
• JavaScript SDK: The application submits prompts and parameters using the textTo3d method, and the model address is obtained asynchronously. Developers need to handle situations such as timeouts, failures, invalid results, and expired addresses.
• Unity editor tool: Open Text-To-3D from Unity’s window menu, enter the prompt to generate and preview the result; once satisfied, save it to the project’s asset directory.
• Unity script interface: It allows for the invocation of generation methods during execution, enables the setting of callbacks for success, errors, and progress, and facilitates the retrieval of metadata such as messages, changed values, and generation time.
• Web compatibility: GLB can be used directly with WebGL and browser-based 3D viewers, making it suitable for product displays, interactive prototypes, and lightweight games. The number of faces in the model, the size of the textures, and the loading time still require optimization by developers.

Input, Output, and Files

JavaScript integration process

1. Ensure that the project uses a supported JavaScript or TypeScript toolchain, and install the MeshifAI package from npm.
2. Import the SDK on servers or controlled runtime environments; do not expose keys and calling logic whose security status has not been verified to visitors.
3. First, call textTo3d to generate a texture-free model, and use a brief and clear description of the object to verify the process.
4. Read the returned model address, download or load the GLB, and implement error handling for network failures, timeouts, and empty results.
5. Adjust parameters such as variance as needed, and record the prompts, parameters, SDK version, and generation time.
6. Only after the texture-free process becomes stable should you try the textured option, and you need to have a retry strategy in place for longer generation times and failures.
7. Place the model in a web-based 3D viewer or in the subsequent asset processing pipeline, to check the grid, scale, materials, UVs, number of faces, and loading performance.

Usage process of the Unity editor

1. Import the tool according to the Unity installation instructions, and open the Text-To-3D window of MeshifAI in the project.
2. Enter the desired model description and click Generate; once the processing is complete, preview it in the editor.
3. Rotate, scale, and view the grid shape; if you’re not satisfied, make the necessary adjustments and regenerate it – there is no need to save each result.
4. After confirmation, save the file; the tool will place the GLB file in the model directory corresponding to GeneratedModels.
5. Drag assets into the scene, and set their scale, material, collision properties, lighting, LOD, as well as whether they are static or dynamic.
6. Run performance tests on the game, and check for model licenses, brand elements, and any inappropriate content before release.

Integration of Unity scripts

• Use the methods of the GenerateModel class to submit prompts, and register a success callback to add the returned GameObject to the scene.
• Display understandable error messages for failed callbacks; do not allow network issues to block the game’s main thread or cause infinite retries.
• Use the status callback to update the progress bar, but do not treat the estimated progress as the exact completion time.
• Read the prompts, variation values, and generation time from ModelGenerationData to facilitate the debugging and auditing of user-generated content.
• If end-users are allowed to enter prompts, content filtering, rate limiting, cost control, caching, and a manual reporting mechanism should be implemented.

Suitable for users and scenarios

• Independent game developers: Quickly create items, simple environmental objects, and placeholder models to accelerate the development of gameplay prototypes.
• Web developers: Dynamically load GLB in product displays, educational pages, and interactive experiences.
• AR/VR team: Quickly explore spatial objects using text, and then hand them over to professional artists for refinement.
• Teachers and students: Create simple assets for concept explanations, project demonstrations, and 3D programming courses.
• Product and creative team: Verify the shape, composition, and scene orientation before investing time in modeling.
• It is not suitable for directly generating final assets that require strict dimensions, tolerances, engineering certification, medical precision, or AAA-grade topology.

Platforms and components

Price and quantity used

The standard subscription price, free usage quota, points deduction, concurrency limits, overage fees, enterprise plans, or refund policies are not specified. Although the SDK can be installed for free, model generation relies on remote services; therefore, it is necessary to verify the account details, keys, usage amounts, and commercial terms before deployment.

SDK, GitHub, and open-source status

On npm, JavaScript packages are provided by publishers corresponding to the documentation’s owners, and the package metadata indicates a MIT license. This license applies to the code in the SDK packages, permitting its use, modification, and distribution under the specified terms; however, it does not apply to cloud-based models, generation services, websites, Unity components, or generated assets.

There is no official GitHub repository for the complete product, nor is there a single license document that covers all components. Developers should retain the MIT copyright and license text in the installation versions, and they must also check the terms related to the Unity tools, dependencies, and the server side.

Privacy, security, copyright, and commercial use

• No public privacy policy, terms of service, security guidelines, list of subcontractors, location of data, retention period, or deletion procedures were identified.
• Text prompts are sent to a remote generation service, and the address of the model returned may also be hosted in the cloud. Do not submit client secrets, unreleased character designs, personal information, or project materials that cannot be transmitted externally.
• The public documents do not indicate whether the prompting or generation models are used for training, nor do they reveal the supplier of these models. Business teams should obtain confirmation of the data processing and training policies before using them.
• The document describes the output as suitable for commercial use, but it does not contain provisions regarding licensing for the generated assets, exclusivity rights, similar content, or handling of trademarks and infringement issues. A clear commercial license must be obtained before official release.
• The generated models may resemble existing characters, brands, products, or protected designs; therefore, users need to conduct reviews regarding copyright, trademarks, appearance, and content.
• Generation at runtime introduces costs, delays, failure risks, and the possibility of inappropriate content; therefore, calls should be restricted on the server side, auditing should be carried out, and manual intervention should be provided.

Advantages and usage limitations

• Advantages: The path from text to GLB is straightforward, and the resulting output can be used in web pages, as well as in Blender, Unity, Unreal, and AR/VR tools.
• Advantages: It offers both JavaScript and Unity interfaces, making it easy for web and game developers to integrate it into their projects.
• Advantages: The Unity editor allows for previewing the result before and after creation; if it isn’t satisfactory, one can try again without affecting the project’s assets.
• Limitations: Texture modeling remains highly experimental; its generation speed and stability are lower than those of the texture-free mode.
• Limitations: Public documents do not provide any guarantees regarding grid quality, number of faces, topology, UVs, skeletons, animation, or reference data.
• Limitations: Issues related to price, quotas, API SLA, privacy, commercial use, and insufficient information on refunds mean that it cannot be used as a key component for production systems.
• Limitations: AI-generated models usually require manual adjustment of proportions, collisions, topology, materials, and performance.

Summary

MeshifAI is suitable for JavaScript and Unity developers who need to generate GLB prototypes from text quickly, and it is especially useful for early testing in the fields of gaming, WebGL, and AR/VR. Before launching it officially, it is necessary to test the process without textures first, assess the experimental texture outputs, and address issues related to pricing, service stability, data processing, and commercial licensing.