Invoke
Refine generated images on a self-hosted canvas with boards and reusable workflows.
- Starting price
- Free self-hosted application
- Free access
- Free Apache-2.0 software
- Billing unit
- Own runtime; API models separate
- Platform
- Local server with browser canvas
- Verified
- 20 Oct 2026
Overview
What you provide
- Provide a prompt, source image and canvas mask, then select a compatible generation model.
What you get
- Invoke produces iterative image edits through a unified canvas, galleries, boards and reusable workflows.
Choosing this tool
Invoke is useful when a visual workflow revolves around inpainting, outpainting and organizing iterations on boards. ComfyUI gives more explicit graph control; DiffusionBee is a simpler Mac entry point. The self-hosted application and the vendor’s commercial service should be evaluated as separate deployment choices.
A decision worked through
A local SDXL workflow and an API-only GPT Image workflow can use the same interface but have different bills. Free Invoke software removes neither local GPU constraints nor external API usage charges; record the model route before estimating an editing batch.
Invoke alternatives for this task
For Invoke, compare ComfyUI, DiffusionBee. Match the input, delivery format and operating responsibilities before comparing price.
| Product | Price / unit | Free allowance | Input → output | Key limit |
|---|---|---|---|---|
| InvokeRead profileOfficial source ↗ | Apache-2.0 self-hosted: free | No software generation allowance | Canvas edits become revised images. | API-only models add provider costs. |
| ComfyUIRead profileOfficial source ↗ | GPL-3.0 runtime: free | Local generations use own hardware | Node graphs become image outputs. | Custom nodes require review. |
| DiffusionBeeRead profileOfficial source ↗ | AGPL-3.0 GUI: free | No cloud-image quota in local mode | Prompts/masks become edited pictures. | Mac and model compatibility required. |
Setup & workflow
- Install the official Invoke Launcher on compatible hardware.
- Import a supported model and create a board for the image project.
- Iterate with masks or workflow nodes and retain the selected revision metadata.
Requirements & installation
- Use the official launcher at https://github.com/invoke-ai/launcher/releases/latest and install model weights appropriate for the machine.
Limits & review
- Compatibility depends on model size and hardware; no universal VRAM minimum is asserted here. The README labels several models API Only, so selecting them can move data and cost outside a local-only workflow.
Your part
- Review masked boundaries and preserve the settings used for each approved revision.
Sources & verification4
This profile is based on official sources, not a hands-on product test.
Vendor descriptions and demos document advertised features. Editorial guidance is based on these sources.
- Official launcher, canvas and API-only modelsgithub.com
- Invoke official referencegithub.com
- ComfyUI official referencegithub.com
- DiffusionBee official referencegithub.com
