QuickMaker QuickMaker

Documentation

Installation

  • Download the ZIP file from the Addon page.
  • In Blender Preferences open addons, choose Install, select the downloaded ZIP file (do not unzip), confirm, then enable the checkbox and sign in with your account.
  • When updating uninstall the previous version, close Blender, reopen it, and install the new ZIP.
  • The addon alerts you if a new version is available for installation.

Getting Started

  • Create an account on the QuickMaker site and confirm your email.
  • Download the Blender addon from the Account page after logging in.
  • Install the addon in Blender, then sign in with the same credentials in the preferences window.

Preferences

  • QuickMaker Menu: Ctrl + Shift + A opens the QuickMaker tools. The hotkey is customizable.
  • Size: Adjusts the font size of in-app messages.

Tools

  • Prompt: Type the description for your generation.
  • Prompt Suggestion: Sends your existing prompt to an LLM for refinement, free of charge. Use it as a drafting assistant, then edit the result before generating.
  • Prompt Templates: Inserts a hand-selected starter prompt. Templates are starting points, not locked presets.
  • Generate 2D: Generates an image from the prompt, selected image plane, QuickMaker Memory, or Moodboard depending on the active model and settings.
  • Generate Projection: Generates a refined 2D image from the active viewport and automatically applies it as a texture projection on top of the existing asset. Move the created Projector empty to adjust the placement, you can also change the Mask strength in the shader.
  • Generate Depth: Generates a depth map from the selected image and builds a displaced mesh with a Geometry Nodes modifier. Apply the modifiers when you want to freeze the result.
  • Generate 3D: Generates a 3D model from the prompt and selected image reference. Supported models expose texture, quality, polygon, or Multi View controls in the side panel.
  • Generate Combined: Runs a 2D generation first and immediately sends that generated image into a 3D generation. The intermediate 2D image is not imported unless the flow fails before the 3D step.
  • Screenshot: Draw a region to capture the viewport. The capture is stored on disk or in QuickMaker memory depending on the Use QuickMaker Memory setting.
  • Use QuickMaker Memory: Stores screenshots as a reusable hidden reference instead of cluttering the scene. Disable it when you want screenshots imported as editable image planes.
  • Paste Image: Imports an image directly from the system clipboard.
  • Import Images: Imports one or more image files as Blender image planes.
  • Draw: Creates a Grease Pencil object and enters drawing mode for quick sketches or markup.
  • Make Alpha: Converts an image plane into an alpha card with shader controls. It works best with clean black or white backgrounds.

Workflows

Image References

  • For single-reference generation, select the image plane you want to use and make it the active object.
  • For multi-reference 2D generation, select supporting references first and select the main working image last so it is active.
  • QuickMaker uploads selected references before submitting the job. If upload rate limits are hit, wait before retrying.

Moodboard

  • Moodboard is available on supported 2D reference models.
  • Add up to 4 image planes to the Moodboard. Duplicate images are rejected.
  • Moodboard images persist while switching models, but only models with multi-reference support use them.
  • The prompt is automatically annotated so the model treats Moodboard images as style, material, mood, and design references rather than exact layout instructions.

Multi View 3D

  • Multi View is available on supported 3D models such as Tripo H3.1 and Hunyuan V3/Pro.
  • Enable Multi View in the 3D model controls.
  • Select exactly 3 image objects in this order: Back first, Left second, Front last. The Front image must be the active object.
  • QuickMaker then sends the images to the provider as Front, Left, Back. If the selection order is wrong, the job can succeed but produce strange geometry.
  • Each Multi View object must contain exactly one image texture, and all 3 images must be different files.

Video

  • Video generation requires an image reference; pure text-to-video is not currently exposed.
  • For normal image-to-video, select one image plane and generate.
  • For first/last-frame video, select the start frame first and the end frame last so the end frame is active.
  • Duration, resolution, and aspect-ratio controls are shown only for models that support them.

Models

2D

  • Nano Banana: General image generation and editing.
  • Nano Banana 2: General image generation and editing with strong value.
  • Nano Banana Pro: Premium image generation and editing.
  • GPT Image 2: Image generation and editing with selectable output size and quality.
  • Flux Kontext: Prompt and reference-based image editing.
  • Flux 2 Flash: Fast image generation and editing with selectable resolution.
  • Topaz Upscale: Image upscaling. Cost changes with output megapixels and upscale factor.
  • SeedVR2: Image upscaling with factor or target-resolution modes.

3D

  • Tripo P1: Premium single-image 3D generation. Polygon Limit 0 uses the model default.
  • Tripo H3.1: Single-image or Multi View 3D generation. Polygon Limit 0 uses the model default; custom values are clamped to 1,000 through 2,000,000.
  • Hunyuan V3: Image-to-3D generation with texture toggle and Multi View support.
  • Hunyuan Rapid: Faster image-to-3D generation. Uses the texture toggle but does not expose custom polygon count.
  • Hunyuan Pro: Higher-detail image-to-3D generation with texture toggle and Multi View support. Polygon Limit 0 uses the model default; custom values are clamped to 40,000 through 1,500,000.
  • Meshy V6: Image-to-3D generation with polygon, pose, and rigging controls.
  • Trellis 2: 3D generation with resolution, polygon, and texture-size controls. Multi View is disabled for this model.
  • Trellis 2 Retexture: Retextures the selected mesh from a reference image. Select the mesh to export and provide the reference image required by the workflow.

Upres

  • Ideogram Upscale: Upscales a selected low-resolution image and requires a descriptive prompt plus an image reference.

Video

  • Veo: Premium image-to-video and first/last-frame video generation.
  • Kling: Premium image-to-video generation.
  • Kling 3 Pro: Premium image-to-video generation.
  • Seedance 2: Image-to-video generation.
  • Seedance 2 Fast: Faster image-to-video generation.

Tips

  • Use detailed prompts for the best results.
  • If you hit rate limits let us know so we can adjust them.
  • To "freeze" Projection or Depth, simply apply all the modifiers on the mesh.
  • If you want to use references for your 2D image generation keep in mind that QuickMaker will upload as many images as you select. Your working image should be the last thing you select, so it is active.
  • Please wait until the Screenshot tool has fully launched before dragging to take a screenshot, in some cases it can take a moment to begin. If for any reason you change your mind and want to exit the screenshot tool, give it 10 seconds for the timeout to complete.
  • Video generation only works from selected images, not pure prompt, so make sure the correct image object is active.
  • With AI your prompt is extremely important, so we built the Prompt Suggestion tool to assist artists with what we believe is the best format to prompt for great results. You can use it to see a specific example, but generally we suggest the following format: Imagine yourself as a senior { } artist/designer, then include the details of what you want to create ie: medieval knight in castle ruins, and then finally add the technical details such as the camera angle, lens type, lighting information and so on.

Troubleshooting

  • Generation failed: Confirm your account has enough credits, then check Blender's Window -> System Console for the provider or upload error. Failed provider jobs are refunded automatically.
  • Upload rate limited: Too many reference uploads were requested in a short period. Wait before retrying. If a multi-image upload is rate limited after an upload session starts, QuickMaker refunds the unbound upload charge.
  • Polygon limit failed: Use 0 for default. Tripo H3.1 supports up to 2,000,000; Hunyuan V3/Pro supports up to 1,500,000.
  • Multi View looks wrong: Re-check selection order. Select Back first, Left second, Front last, and make sure Front is active.
  • Screenshot does not work: On macOS, allow Blender screen recording access. On managed machines, ask IT to allow Blender outbound HTTPS traffic.
  • Video or 3D jobs take longer: Keep working in Blender while QuickMaker polls the job and imports the result when it is ready.
  • Selected references behave unexpectedly: Select supporting references first and the main working image last so it is the active object.

Support

Changelog

Version 1.1

  • First and Last Frame video generation added for supported video models. Select a start image and an active end image to generate transitions between frames.
  • Expanded video model support, including Veo, Kling, Kling 3 Pro, Seedance 2, and Seedance 2 Fast workflows.
  • Improved reference image handling across 2D, 3D, and video generation. QuickMaker now handles selected references, active working images, and model-specific reference fields more consistently.
  • Moodboard workflow added for supported 2D reference models. Moodboard images persist when switching models and are used automatically by models that support them.
  • Moodboard panel now shows the current image count and gives a clear popup error when adding a duplicate image.
  • New and expanded model support, including Nano Banana Pro, GPT Image 2, Flux 2 Flash, Topaz Upscale, Tripo P1, and newer video models.
  • Added model-specific controls for resolution, quality, aspect ratio, upscale settings, polygon limits, texture options, duration, and video resolution where supported.
  • Multi View and 3D reference workflows improved so supported 3D models receive the correct image order and model-specific fields.

Version 1.0

  • Nano Banana 2 model added.
  • LLM based Prompt Suggestions now have multiple other free models to use as fallbacks to ensure maximum uptime.
  • Generate Projection.
  • Generate Depth.

Version 0.9

  • Additional Prompt Templates.

Version 0.8

  • Backend Update.

Version 0.7

  • Backend Update.

Version 0.6

  • Prompt Templates.
  • Alert the user to save the file before generating assets if it's not saved.
  • Go into material preview mode if user is not in shaded view when generating to ensure a proper preview.
  • New panel for the Prompt box so the user can see the entire prompt at once and have a bigger text area to type in.
  • Improved LLM suggestions

Version 0.5

  • Backend Update.
  • Nano Banana Pro model added.

Version 0.4

  • Backend Update.

Version 0.3

  • Prompt Suggestion: LLM suggestion for your prompt.
  • Kling and Veo Video models added.

Version 0.2

  • Screenshot: captures a region with the system tool, saves to disk, and confirms completion in the UI.
  • Use QuickMaker Memory: If enabled it saves Screenshots to the QuickMaker Memory, if disabled it will import the Screenshot immediately into Blender.
  • QuickMaker Menu hotkey defaults to Ctrl + Shift + A.
  • QuickMaker performs all processes in background so you are free to work in the viewport while assets generate.
  • Imported plane shader uses an emissive setup with roughness 1 and specular 0.
  • UI display text moved to the lower third of the screen.
  • Addon automatically checks for updates.
  • Generate Combined: chains 2D and 3D generation for faster workflows.
  • Make Alpha: builds alpha cards from image planes with ramp control.
  • Hunyuan 3D model added.

Version 0.1

  • Initial launch.
  • Nano Banana, Tripo 3D models added.