Blender is a free and open-source 3D creation suite developed by the Blender Foundation. It provides tools for modeling, rigging, animation, simulation, rendering, compositing, motion tracking, video editing and game creation. Blender is distributed under the GNU General Public License.
Higgsfield for Blender is an AI add-on that brings Higgsfield's generation tools into Blender, generating and importing 3D scenes, meshes, rigs, textures, images, and video. It adds a floating viewport bar with tabs for scene building, 3D models, character animation, images, video, cameras, and assets; results can be inserted into the open scene as editable geometry, planes, materials, rigs, keyframes, or video outputs. Scene Builder creates editable objects, layouts, and lighting from a prompt. The character-animation workflow produces a fitted weighted rig with timeline keyframes. An optional Higgsfield Bridge MCP endpoint at https://bridge.higgsfield.ai/mcp connects an external AI agent such as Claude to the add-on for operations including building blockouts, generating meshes at the 3D cursor, and requesting Seedance renders in the open Blender scene. Generation runs on Higgsfield's servers, requires an internet connection and a signed-in Higgsfield account, uses the account's existing Higgsfield credits, and supports Blender 4.2 through 5.1.
Searchable transcript of How I Make Cinematic AI Movie & Product Shots (Blender + AI Workflow) — ManuAGI - AutoGPT Tutorials (06:40). Search for a phrase, then click its timestamp to jump straight to that moment in the video.
Captions sourced from the original video on YouTube, published by ManuAGI - AutoGPT Tutorials. The video, its captions and all related intellectual property remain the property of their respective owners; AINotes claims no ownership. Provided for research, accessibility and search — see the Transcript Notice and Copyright Policy.
00:05 This looks like a finished cinematic game cutscene with a character firing a massive weapon, the muzzle flash filling the frame, and the character dropping back onto the floor. But this shot started as a simple gray blockout inside Blender. The environment, character, weapon, camera movement, and final fall were planned before the finished look was generated.
00:28 That is the workflow we're going to build with Higsfield's Blender plugin. We'll create the 3D scene, animate the character, control the camera from a phone, and use the blockout to generate the finished version. [music] The point is to decide the shot first, then let the video model add the detail. This video is sponsored by Higsfield. Let's get into the workflow.
00:54 Higsfield Blender MCP puts an AI agent directly inside Blender. You can describe a scene, generate a 3D model, place it into the scene, create camera movement, add animation, and turn the whole blockout into a video reference without building everything manually. The important part is that we are not starting with a random text prompt and hoping the camera does something useful.
01:19 We are going to build the structure of the shot first. Lock the movement and then use Higsfield to generate the final video around it. Let's start with the setup. First, download the Higsfield Blender plugin from the official page. It comes as a zipped file and you can drag that file onto an open Blender window to install it. After that, sign into your Higsfield account.
01:42 Once it is active, the Higsfield bar appears over the Blender viewport. This is where the main workflow happens. You have tabs for scene builder, 3D model, character animation, image, video, camera, and asset. The prompt bar lets you describe what you want to create, choose the relevant model, adjust settings like resolution and frame rate, and see the credit cost before generating.
02:08 There is also a bridge connection that lets an AI agent work inside your open Blender scene. Copy the Higsfield Bridge URL. Open your agents connector settings. Create a custom connector. Name it Higsfield Bridge and paste in the URL. Once you sign in, you can ask the agent to build directly in Blender. For this example, I'm going to create a simple product commercial around a futuristic camera.
02:38 I'll start by asking the scene builder to create a small product stage. I want a dark studio room, a flat platform in the center, a large curved background, soft lighting from the left, a stronger rim light from the right, and enough empty space around the platform for the camera to move. The goal here is not to create a finished environment. We only need the important structure, the floor, the background, the lights, and the main area where the product will sit.
03:08 This is the first blockout. It gives us the basic layout before we spend time on details. Now, I'll move to the 3D model tab and generate the hero object. I'll describe a compact futuristic camera with a matte black body, a large circular lens, simple mechanical details, and a clean product design shape. The model is generated at the 3D cursor and comes into Blender as editable geometry.
03:36 That means I can move it, scale it, rotate it, or place it directly on the platform. I'm keeping the object simple on purpose. A single product is easier to control than a scene with 10 different characters and props, especially when the main goal is to lock the camera movement. Once the model is ready, I'll open the asset tab. This is where generated 3D results, images, and videos are stored.
04:02 I can find the model again, place it into the scene, and keep the project organized without searching through separate folders every time. Next, I'll use the image tab to create a visual reference for the product surface. This could be a label, a logo treatment, a material idea, or a background texture. The image result can be inserted into the scene as a plane, or used as part of a material.
04:31 So, the final object has a clear visual detail when the camera moves closer. At this point, we have a studio, lighting, a product model, and a basic visual treatment. We are not trying to make the Blender scene look like the final commercial. We are building the structure that the video model needs to follow. The camera tab lets you control the Blender camera with your phone.
04:52 Scan the QR code that says, "Scan with your phone to drive this camera. Connect the device and move the phone to control the camera inside the scene." The panel includes camera, walking, record, recenter, home, look, zoom, and speed controls. For this demo, the phone connection and a short recorded camera movement are enough to show the feature. The camera path is saved directly into the Blender timeline, ready to be used in the animation.
05:24 Now, we move from the camera to the subject inside the scene. Open the character animation tab and select the generated character. This is where Higsfield adds a usable rig and animation to the character instead of leaving it as a static 3D model. Choose the movement we want, such as walking, turning, reaching toward the product or reacting to something in the scene.
05:48 Higsfield fits the character with a weighted rig and places the animation as key frames directly on the Blender timeline. Now the camera movement is ready so the scene can become an animated shot. Preview the camera view and check that the movement starts and ends correctly. The materials and lighting can remain simple because this render is only being used as a reference.
06:09 This gives us a video showing the exact camera movement, framing and timing. Once the first shot works, I do not need to rebuild everything from zero. The same camera path can be reused with a different product, a different background, or a completely different visual style. That is what Higsfield Blender MCP helps with. It gives you a way to plan the scene and camera before the final generation.