Overview of OpenBlender

I have summarized the overview of OpenBlender.

1. OpenBlender

“OpenBlender” is an add-on for integrating generative AI features into Blender. While creating 3D scenes in Blender, you can handle image generation, video generation, 3D model generation, HDRI generation, rig generation, AI chat, and external operations via MCP within the same screen.

By using “OpenBlender,” you can use Blender not just as a 3D production tool, but also as a composition control tool for generative AI, a material generation tool, and an operation target for AI agents.

The main features are as follows:

A key feature is that you can build a generative AI workflow centered on Blender, rather than performing image or video generation outside of Blender and only bringing the results into it.

2. Main features of OpenBlender

2-1. Image Generation

In “OpenBlender,” in addition to “TXT to IMG,” which generates images from text, you can also use “IMG to IMG,” which generates new images based on existing ones. What is particularly interesting is that you can capture the Blender viewport and pass it to IMG to IMG.

For example, you can create a simple 3D scene in Blender, determine the camera position, object placement, lighting, and depth, and then generate an image.
When specifying composition using only text prompts, it can sometimes be difficult to consistently control camera angles, depth, and the positional relationships between objects.

On the other hand, if you create a rough composition in Blender first, you can determine the screen design in quite a bit of detail before passing it to the generative AI.

For example, the flow is as follows:

Since you can fix the composition in 3D before generating the image, it is convenient when you want to finely control the background, characters, buildings, and camera position.

2-2. Video Generation

With “TXT / IMG / VID to VID” in “OpenBlender,” you can handle video generation using text, images, or videos as input. Not only can you generate videos from text alone, but because you can generate videos based on images or videos, you can use the composition and camera work created in Blender as a guide for generative AI video.

For example, you could create camera animation in Blender and use its movement and structure to generate a video.
Generative AI for video can sometimes result in unstable composition and movement with just prompts. Therefore, by using Blender as a layout tool, you can pre-design the camera position, depth, and movement direction.

It is well-suited for a workflow of “creating the foundation of movement in Blender and adding visual expression with generative AI.”

2-3. 3D Model Generation from Images

"IMG to 3D" allows you to input an image and generate a 3D model in GLB format via "Trellis.2", which can then be imported into Blender with textures. Since the generated model is loaded in GLB format, you can place it, edit it, or adjust its materials directly within Blender.

The general workflow is as follows.

You can use this to turn characters or props created via image generation into 3D objects and place them in your Blender scene. Of course, it is more realistic to use the generated 3D models as rough drafts or background props rather than as finished products. Even so, having a workflow that connects image generation to 3D production directly within Blender is convenient.

2-4. Text to HDRI Generation

"TXT to HDRI" uses "FLUX.2-Klein" and a LoRA for HDRI to generate a 360-degree equirectangular panoramic HDRI environment map from a text prompt. The generated HDRI can be applied to the World environment in Blender. It is useful when you want to quickly test the atmosphere of a background or lighting.

For example, you can specify environments like the following via text:

HDRI significantly influences the atmosphere of a 3D scene. Normally, you would need to search for HDRI assets or create them using external tools, but with OpenBlender, you can generate them from text and apply them to Blender. This seems useful for quickly exploring the direction of your lighting and background.

2-5. Text to Rig with Motion Generation

"TXT to RIG" allows you to specify motion via text to generate a generic humanoid dummy rig with animation. This is not a feature for generating character appearances or textures, but rather a tool to be used as a starting point for motion.

For example, you specify actions via text like the following:

Based on the specified action, you can generate a humanoid skeleton animation and load it into Blender. It is best to view this as a feature for creating a motion draft rather than generating a character's appearance. If you consider retargeting to your own character, this can shorten the initial stages of animation production.

2-6. Remix and LoRA Support

"OpenBlender" also includes a "Remix" feature. With "Remix", you can generate variations based on an existing image. This can be used when you want to create alternative versions based on render results, generated images, textures, or HDRIs. Since you can check the generation results within Blender and try out variations immediately, it makes trial and error in image generation much easier.

It also supports "LoRA". By using "LoRA", you can generate images that reflect specific styles, characters, art styles, or textures. By incorporating LoRA into the ComfyUI workflow, you can reflect your preferred expressions in image and video generation from "OpenBlender".

2-7. AI Chat

"OpenBlender" also includes an "AI Chat" feature. You can display an AI assistant within Blender using "OpenRouter" or "LM Studio". The configuration uses "OpenRouter" for cloud LLMs and "LM Studio" for local LLMs. For features that use image analysis, vision-capable models such as GPT-4o, Claude, or Kimi K2.5 are intended.

AI Chat can be used not just as a conversation partner, but as a tool to assist with generation tasks within Blender. For example, you can show the AI three keyframe images and have it create a prompt for video generation.

generate a video prompt from keyframes images

By making a request in this way, you can create video prompts based on the content of an image, including scenes, camera work, lighting, atmosphere, and transitions. In image and video generation, creating the prompt itself is time-consuming. With "OpenBlender," it is convenient to be able to perform that prompt creation in a context close to your work within Blender.

2-8. MCP Server

OpenBlender also has an "MCP Server" feature. It starts an HTTP/SSE server on the Blender side, allowing external MCP clients to control Blender.

The default connection destination is in the following format.

http://localhost:9876/sse

Using the "MCP Server" allows you to call Blender scene operations, object creation, material editing, animation settings, rendering, and viewport operations from an AI agent.

For example, imagine giving instructions like the following from an MCP-compatible client.

If you connect with MCP-compatible tools such as "Claude Desktop," "Claude Code," "Cursor," or "OpenCode," you can aim for a workflow where you build a Blender scene while giving instructions via chat. It is convenient to be able to handle "Blender MCP"-based workflows all within OpenBlender.

3. System Requirements

"OpenBlender" assumes a fairly heavy local generative AI environment.

As for the system requirements, the following configuration is generally recommended.

In particular, if you try to use all features such as image generation, video generation, 3D generation, and HDRI generation, the model capacity will become quite large. It is best to first enable only the features you want to use and install only the necessary models.

If you are already using ComfyUI, you can also consider a configuration that shares an existing model directory. Be aware that downloading duplicate models will consume a significant amount of storage.

4. Price and How to Purchase

"OpenBlender" is sold on "Gumroad." The price is "€12.99+." In Japanese yen, depending on the exchange rate, it is approximately 2,400 yen (as of the end of May 2026).

The "+" notation on Gumroad can mean that you can pay more than the minimum price at your discretion. When purchasing, please check the current price, included files, updates, and license terms on the Gumroad product page.