Technology has always existed to make tasks easier for humans, and advancements in technology keep getting faster and smarter. In the 21st century, groundbreaking achievements have been made in the field of artificial intelligence, such as the development of machine learning processes and Chatbots like ChatGPT.
One field where AI tools have gradually begun to enter is the 3D marketplace. Normally, a skilled operator would have to utilize 3D modeling software, such as Blender, and exert a great deal of time and effort in order to create complex 3D models. With the aid of an AI tool, the entire design process is not only faster and easier for professionals but also lowers the learning curve for beginners – making the creative process and bringing their ideas to life much easier.
With that said, finding the AI tool that specifically suits your purposes can be a time-consuming task. Therefore, we have compiled this list of some of the best AI tools available online that can be used for 3D model generation.
Criteria
In order to gather various recommendations that are suitable for a variety of users, we kept in mind the following considerations while compiling this list:
- Popular: Suffice to say, positive user feedback is a must. Many of these AI tools are still being refined, but that doesn’t make them any less fun to experiment with!
- User Friendly: The primary reason to look for an AI is to make the job easy, thus the software should be easy to use and navigate.
- Versatile: It would be preferable to find an AI that’s capable of producing a variety of models and can be used for different purposes compared to one that’s only usable for a specific purpose.
- Affordable: Good things may often come at a price, but it doesn’t leave a dent in your wallet. Even better if it’s free.
This list will contain a variety of tools, including text-to-3D model generators that produce a model based on text prompts, and image-to-3D generators that do the same task but based on user-uploaded image(s). These can be browser-based tools, downloadable software, or even source code.
Genie
Here’s one of the hottest AI tools at the moment, AI Genie by Luma. This tool can easily create detailed, high-resolution textured models that are downloadable in various formats. While other tools may require some technical knowledge to install and use, Genie can be run in the browser, Discord, or through the Genie iOS app, making the entire process as straightforward as it gets.
If you’ve never used an AI tool, this may seem more like magic than technology – only you’ll have to sign up to use it first. The creation of a model begins with a text prompt where you must describe the object you want followed by other desirable characteristics or styles separated by a comma. After processing, the Genie will show you four low-quality versions of the model that you can select from for further refinement and download.
The number of samples one can generate is unlimited and completely free, so if you’re not satisfied with the results you can try generating new variations or getting more specific with the text prompt. It’s worth keeping in mind that like all tools of this kind, it’s not going to be able to compete with the quality of design that a human can produce. You may end up with lots of refinement to do on the model and various mesh oddities.
For best results, take into account that Genie is specialized in generating individual objects and is not yet capable of creating complete scenes or environments. Transparent materials are also a challenge and may appear correctly in the preview but give errors after being exported in mesh formats.
If you’re interested in exploring this tool to its full potential, we recommend joining the Discord where you can get access to guides, tips, and additional features like a refinement and texturing progress bar.
- Price: Free
- Key feature: Text to 3D model
- Export formats: STL, BLEND, OBJ, FBX, GLTF, USDZ
- Platforms: Browser-based, Discord, iOS application
- Site: Luma AI
Meshy
Meshy is another powerful tool with incredible features that include generating 3D models and textures from text input. To use the tool, you’ll just need to head to the web application where you’ll find both Beta and Alpha versions after signing up.
In addition to a text prompt you can use to describe the model you want, there’s also a separate “Negative Prompt” that you can use to describe characteristics you’d like the AI to exclude from the model. There’s also a few basic art styles to choose from. Like other tools, the initial generation will give you four models to choose from that you can then refine further and download.
The AI texturing tool can be integrated with Unity Game Engine or Blender through an API to help you build your vision faster than ever. It’s also possible to upload a model without texture to the web application and have it generate a PBR texture map for the model right in the browser.
While Meshy promises that even someone who has never had contact with 3D models can generate their assets easily, it’s worth checking out the available video tutorials to understand how to best utilize the tool. Other videos on MeshyAI‘s YouTube channel include tutorials for making Minecraft-like voxel-style models.
For those just looking to create the odd model or two, the free plan should suffice: You get 200 credits that are reset and renewed every month. Each request to the prompt consumes 5 credits to generate the initial models and another 20 for each round of further refinement. All generated assets are shared publicly, but if you opt for a subscription plan, the generated assets will be owned by you. Plans start at $16 a month for a yearly subscription and that gets you 1,000 credits per month and a faster server, among other benefits.
- Price: Limited free plan; annual and monthly subscriptions starting at $16 per month
- Key features: Text to texture, image to texture, text to 3D model, image to 3D model
- Export formats: STL, BLEND, OBJ, FBX, GLB, USDZ
- Platforms: Browser-based, Discord, API interface
- Site: Meshy
Point-E
Made by OpenAI, the developers of ChatGPT, Point-E is a free and open-source AI for generating all kinds of 3D models. It uses input text to search and compile relevant images. The set of images is used to run a point diffusion, producing a 3D point cloud. The point cloud model can then be used to create a 3D mesh file and converted to other 3D formats. It can also produce a model directly from a set of user-supplied images.
The code for this tool is available on GitHub, however, there is no installer released. So, users have to build the software themselves. Users with programming knowledge, specifically Python, can build an executable file for the code, or run it on Jupyter Notebook.
For beginners, access to this AI is available on community platforms like huggingface, although it can sometimes be time-consuming to find a properly running space. An alternative is to run it online using Google Colab. There are a handful of tutorials on YouTube to help out, like the one posted by CGMatter. A similar tool, Shape-E, to directly generate 3D models is in the early experimental phase.
- Price: Free
- Key features: Text to image, image to 3D model, point cloud generation
- Export format: Point cloud
- Platforms: Comes as source code, operable on browsers through Google Colab or software like Jupyter Notebook
- Website: GitHub
Meshcapade
Meshcapade is a fast and simple AI tool that you can run on your browser or a desktop app available for Windows devices. The purpose of this tool is to create 3D avatars, whether for use in Metaverse, VR, or 3D games.
The tool creates a model based on an uploaded image of a person. The model takes the body measurements, then generates a 3D model. You can also create a model directly from measurements if you wish to do so.
The avatar can then be downloaded as an OBJ or FBX file, compatible with 3D game engines such as Unity or Unreal. You can try this tool to generate and export your first model for free, but after that, a credit-based pay scheme is implemented (100 credits are needed to generate a model from measurements). AI-generated measurements cost one token and alignment, which allows for natural movement of the figure, costs five tokens.
- Price: Free trial of 2,000 credits; $5 for 500 credits
- Key feature: Image to 3D avatar generation
- Export formats: OBJ, FBX
- Platforms: Browser-based, desktop app (Windows), API integration
- Website: Meshcapade
CSM
If what you need is to create a model based on a 2D image, this will be a good option to try. With Common Sense Machines (CSM), it’s possible to upload images in PNG, JPG, or JPEG formats and have the algorithm build a 3D model based on the compilation! While the initial model samples and previews may look a bit wonky, the model can turn out quite impressive after some refinement. In addition to using images to generate models, it’s possible to sketch in real-time and, with the help of AI in a text prompt, create a unique 3D model.
The speed of model generation in the free plan can take minutes to hours to refine the final model depending on the complexity and availability in the public server queue. One can generate unlimited free previews, but refining the model takes considerable time and a subscription plan is needed to convert the mesh into high-resolution. Models generated under a free plan will also be available publicly under a Creative Commons license.
CSM can be connected to Discord to generate models from images, and its models are compatible with game engines like Unreal. To learn more about these possibilities, be sure to check out the available tutorials. In any case, this tool greatly expands the possibilities of modeling exactly your vision, even if you just have one rough reference image.
To get an idea of the potential of this tool, just take a look at the community showcase to see beautiful models that can inspire you to get started.
- Price: Limited free plan; annual and monthly subscriptions starting at $20 per month
- Key features: Image, sketch, and text to 3D model; animation
- Export formats: OBJ, FBX, GLB, USDZ
- Platforms: Browser-based, Discord, API interface
- Site: CSM
3DFY
3DFY is an online platform that specializes in creating and training its AI model for 3D model generation with a focus on product design. The company says the quality of its AI-generated models are of similar quality to those created by 3D artists.
Among their various services, our primary interest is 3DFY Prompt, a text-to-3D model generator. Its usage is straightforward: Users first have to choose from a selection of model types, such as swords, cutlery, and sofas. The user can then enter a text prompt to describe the style, colors, and textures to be generated and applied to the model.
The text-to-3D service isn’t free, but a trial of 10 generation credits is offered for users who wish to try it out. Creating a model from a prompt counts as one generation credit. In order to export or download the model in a selected file format, a download credit has to be used.
3DFY offers a tutorial on how to utilize the prompts effectively. Keep an eye out in the future for their image-to-3D service, which is currently in development.
- Price: $5 for 10 generation credits and 1 download, plus various subscriptions starting at $15 per month (billed annually)
- Key feature: Text to 3D model generation based on model library
- Export formats: BLEND, FBX, GLB
- Platform: Browser-based
- Website: 3DFY
Sloyd
Artificial intelligence is a fabulous tool, but it still needs to be trained before it can return really useful results. Sloyd takes a very interesting approach by presenting a wide selection of video game-relevant models that can be parametrically modified by both the AI and the user in dashboards. This approach allows users to focus more on aesthetics than modeling and get a higher-quality mesh to export.
The objective is to quickly create personalized assets, so the reach of the AI is limited to models registered in the database. Sloyd’s AI cannot create people, animals, and scenes, but it can generate textured items such as weapons, buildings, furniture, and props. It’s important to keep in mind that both the AI and the environment itself cannot interpret units of measurement during edits.
To use the tool, simply register for access to the web application. The environment is similar to animation and modeling software with a searchable library of assets to the right along with a few categories you can select from including sci-fi, military, urban, medieval, and furniture. If you don’t find the model you’re looking for, don’t be discouraged; there’s an active community that contributes to the creation of new models.
The free plan includes 30 export credits, but annual subscription plans starting at $15 monthly get you 200 monthly export credits that roll over from month to month. Depending on the selected resolution of model export, the credit cost for model export can vary.
- Price: Limited free plan; annual and monthly subscriptions starting at $15 per month
- Key feature: Text to 3D model generation based on model library
- Export formats: OBJ, GLB
- Platforms: Browser-based, Discord
- Site: Sloyd
Spline AI
Spline is a user-friendly 3D design tool with parametric modeling, 3D sculpting, and allows for real-time collaboration between team members. They have recently introduced AI tools into their features list. With a free membership, users can join the waitlist for access to their text-to-3D model tool, dubbed Spline AI, which generates editable 3D models based only on text prompt descriptions. Text prompts can then be used to add textures and create animations.
While you’re waiting in line for the full release, you could try out Spline’s AI tool called Style Transfer. It relies on text-based prompts to edit existing models and scenes with the goal of minimizing post-processing time. Style Transfer is still in the “experimental” phase and is therefore likely to have bugs, but the developers are actively working on improvements with the help of user feedback.
Style transfer is accessible with a paid membership to Spline Super of $9 per month. In addition to gaining access to a number of extra modeling tools, members also get priority on the Spline AI waitlist.
- Price: Limited free plan (no access to AI tools); annual or monthly subscriptions starting at $9 per month
- Key features: Prompt-based editing tool, text to 3D model generation coming soon
- Export formats: STL, GLB, GLTF, USDZ
- Platforms: Browser-based, desktop app (Windows and MacOS)
- Website: Spline
License: The text of "The Best AI 3D Model Generators in 2024" by All3DP is licensed under a Creative Commons Attribution 4.0 International License.