
What I like most about Meshy is how well it integrates into my AI-driven game development workflow. I use Meshy through its plugin integration with Codex and Claude Code, and it has become an important part of how I create 3D game assets.
My workflow starts with AI generating concept art based on my game ideas. The AI then uses Meshy to turn those images into 3D models. After that, it connects to Blender through MCP to handle rigging, make adjustments, and optimize the models before integrating them directly into my games. Almost the entire process is handled by AI, from the initial concept to the final game implementation.
The biggest advantage for me is that I can manage this entire pipeline through AI agents without constantly switching between different applications or manually handling every step. Meshy's image-to-3D capabilities make it much easier to create custom assets that match the specific style and needs of my games.
I also really appreciate how accessible 3D asset creation has become. Instead of spending hours manually modeling something or searching for existing assets that might not quite fit, I can generate something based on my own concept and have AI refine it further in Blender.
From a productivity and ROI perspective, this is a huge benefit. As an indie game creator, being able to automate so much of the asset creation process allows me to spend more time on actual gameplay development and experimenting with new ideas. Meshy's integration with AI coding agents is what makes it especially valuable to me. Review collected by and hosted on G2.com.
My biggest frustration with Meshy is how API-generated assets are handled. Since I use Meshy primarily through AI agents like Codex and Claude Code, most of my 3D models are generated through the API rather than the web interface. Unfortunately, these assets don't appear in my Meshy profile or personal asset library.
This is disappointing because I would love to see all my creations in one place, regardless of whether they were generated through the website or API. I also can't share my API-generated models with the Meshy community, which feels like a missed opportunity, especially for developers who rely heavily on automated workflows. I think API-generated assets should have the same visibility, management, and community-sharing options as assets created directly on the platform.
Another limitation is the animation library's API functionality. Meshy has a useful collection of animations, but I can't retrieve individual animations directly through the API. Instead, I have to apply an animation to a model and download the resulting asset. For my workflow, where AI agents handle rigging and animation management through Blender MCP, being able to download standalone animations would be much more practical. It would make it easier to reuse animations across multiple characters without unnecessary extra steps.
Overall, I really like Meshy and what it enables, but I feel the API experience still needs to catch up with the web platform. Better integration between API-generated content and the user library, along with more flexible animation API access, would make Meshy significantly more useful for developers building fully automated AI workflows. Review collected by and hosted on G2.com.