# Custom Workflow - 1 Image in ComfyUI-Roboflow You can create whatever workflow you want. This node takes in one image and will output up to 4 images and 4 strings depending on what you are doing. Screenshot 2024-09-25 at 10 36 18 AM ## 1. Install ComfyUI-Roboflow Follow the instructions in [README.md](https://github.com/DareFail/ComfyUI-Roboflow?tab=readme-ov-file#installation) ## 2. Add the Roboflow Custom Workflow - 1 Image Node Right-click Add Node -> Roboflow -> Custom Workflow - 1 Image Screenshot 2024-09-25 at 9 38 30 AM ## 3. Press "Create Workflow" and make your custom workflow Add any of the blocks or edit from a template here. Screenshot 2024-09-25 at 10 11 51 AM ## 4. Click the Response node in your workflow 4 ## 5. Copy the output names you care about 5 ## 6. Paste the output names as keys in your ComfyUI Node Output images are go under output_image1_key, output_image2_key, etc Your input_image1_key is always "image" unless you changed it in Roboflow. 6 ## 7. Go back to Roboflow and click "Deploy Workflow" in the top right Screenshot 2024-09-25 at 9 54 13 AM ## 8. Copy your API Key, workspace name, and workflow id You can copy each part under the python section Your API Key will look like this ***** but if you copy and paste it will be the actual API Key. Screenshot 2024-09-25 at 9 54 04 AM ## 9. Paste the information above into your node Screenshot 2024-09-25 at 10 04 13 AM ## 10. Use your node I linked up a Load Image Node and sent the image out to a Preview Image Node. You can find both under Right Click Add Node -> image -> Load Image Add Node -> image -> Preview Image Screenshot 2024-09-25 at 10 36 18 AM # Repo Info ## License Distributed under the APACHE 2.0 License. See `LICENSE` for more information. ## Contact (feel free to ask questions!) Twitter: [@darefailed](https://twitter.com/darefailed) Youtube: [How to Video coming soon](https://www.youtube.com/@darefail) Project Link: [https://github.com/DareFail/ComfyUI-Roboflow](https://github.com/DareFail/ComfyUI-Roboflow)