Generate or edit images with OpenAI GPT Image from text or camera frames, with save and programmatic control commands.
Windows
How to add the Openai Dall-E Image robot skill
- Load the most recent release of ARC (Get ARC).
- Press the Project tab from the top menu bar in ARC.
- Press Add Robot Skill from the button ribbon bar in ARC.
- Choose the Camera category tab.
- Press the Openai Dall-E Image icon to add the robot skill to your project.
Don't have a robot yet?
Follow the Getting Started Guide to build a robot and use the Openai Dall-E Image robot skill.
How to use the Openai Dall-E Image robot skill
Integrate state-of-the-art image generation capabilities directly into your robot apps and products. OpenAI's GPT Image models create realistic images and art from natural language descriptions, and can edit an existing image to match a prompt.
Have your robot programmatically generate images from speech recognition robot skill descriptions. Or, have the image from the camera sent to OpenAI to be edited with a prompt. Experience the whacky world of AI image generation with your Synthiam-powered robot or webcam.
*Note: OpenAI retired the DALLE 2 and DALLE 3 models on May 12, 2026. This robot skill now uses the GPT Image model family (gpt-image-2, gpt-image-1.5, gpt-image-1, and gpt-image-1-mini), which you select in the configuration menu.
How Does It Work?
The GPT Image models have learned the relationship between images and the text used to describe them. From a text description, they can create original, realistic photos and art, combining concepts, attributes, and styles. They can also edit an existing image (such as a camera frame) to apply a prompt you provide.
Main Window
1) Generate From Description
Prompts a dialog for you to enter a description of the image you would like generated. Be as detailed as possible because anything is possible. The generated image is displayed in the preview box.
2) Modify the Camera Image
Takes an image from the camera device and edits it with a prompt that you provide. A dialog asks how you would like the camera image modified (for example, "Make it look like a watercolor painting"), and OpenAI returns an edited version of that image.
*Note: The GPT Image models edit an image against a prompt; they do not create prompt-less "variations" like the retired DALLE 2 model did. A prompt is therefore required when modifying the camera image.
3) Save Image
Saves the most recently generated or edited image to a file on your computer. A save dialog lets you choose the location and format (PNG, JPEG, or BMP). Generating or modifying an image again replaces the result, so save it before re-running if you want to keep it.
Configuration
1) API Settings
Developer API Key - Enter the API Key from OpenAI's website here. This value will be saved locally on the current user account and used for any OpenAI robot skills. This value will not be distributed with your project.
Image Model - Select which OpenAI image model to use:
- gpt-image-2 - Newest and highest quality, with the largest output sizes. Slower and the most expensive.
- gpt-image-1.5 - A balanced default for quality, speed, and cost.
- gpt-image-1 - The original GPT Image model. A lower-cost option.
- gpt-image-1-mini - The fastest and least expensive option.
Control Commands
There are a few ControlCommand() that allow programmatic usage of this robot skill from other robot skills. For example, send the description from the result of a speech recognition request to be generated as an image. The camera image modification uses the camera device robot skills, so they are automatically queried for the control command. If no camera device name is specified, the first camera device is used.
- GetImageFromText
Example: ControlCommand("OpenAI Dall-e Image", "GetImageFromText", "description of the image to generate");
Generate an image based on the provided description.
- GetImageVariation
Example #1: ControlCommand("OpenAI Dall-e Image", "GetImageVariation", "Make it look like a watercolor painting");
Example #2: ControlCommand("OpenAI Dall-e Image", "GetImageVariation", "Make it look like a watercolor painting", "Camera");
Edit the camera image using the supplied prompt. The first parameter is the prompt describing how to modify the image (required). The optional second parameter is the camera device name.
*Note: When no camera device name is specified, the first camera device will be used.
- SaveImage
Example: ControlCommand("OpenAI Dall-e Image", "SaveImage", "c:\temp\image.png");
Saves the most recently generated or edited image to the specified file path. The file format is chosen from the file extension (.png, .jpg/.jpeg, or .bmp), and the destination folder is created if it does not exist.
API Key
An account is required on the OpenAI website to use the service. Once an account is created, you will have an API key to use in the robot skill.
Visit the OpenAI website and select the API option from the menu.
Press the SIGNUP option to create an account
Once the account is created, press your Username icon to view the drop-down. And select the View API Keys option.
Copy your API key
Paste the API key in the robot skill configuration menu
Control Commands for the Openai Dall-E Image robot skill
There are Control Commands available for this robot skill which allows the skill to be controlled programmatically from scripts or other robot skills. These commands enable you to automate actions, respond to sensor inputs, and integrate the robot skill with other systems or custom interfaces. If you're new to the concept of Control Commands, we have a comprehensive manual available here that explains how to use them, provides examples to get you started and make the most of this powerful feature.
Control Command ManualGetImageFromText", "A goldfish driving a car
Example: controlCommand("OpenAI Dall-e Image", "GetImageFromText", "A goldfish driving a car")
GetImageVariation", "Make it look like a watercolor painting
Example: controlCommand("OpenAI Dall-e Image", "GetImageVariation", "Make it look like a watercolor painting")
SaveImage", "c:\temp\image.png
Example: controlCommand("OpenAI Dall-e Image", "SaveImage", "c:\temp\image.png")
Related Questions
Using Openai Skill With Other Products
AI Generated Results Into Text Of A Certain Region Of Screen...
Openai Dall-E Skill Deprecated
Upgrade to ARC Pro
Elevate your robot's capabilities to the next level with Synthiam ARC Pro, unlocking a world of possibilities in robot programming.
Thanks, @Dave!! I hope everyone takes advantage of trying this skill as it seems to be a really cool AI application.
I'm still learning how to use it effectively. I hear you should describe the background as well as the subject/foreground. You can also choose the style by writing "oil painting" or "Claude Monet".
Here's the description I used to create a picture I had in my mind.
"2 highly detailed ninja robots with a dark and neon Japanese cityscape behind them on a foggy night"
While it wasn't exactly what I wanted I did appreciate the creative output!
It's a very interesting AI. Seems like a lot of fun and a good way to pass some time away. Kinda gets you thinking.
@Athena How do I display the generated image within ARC?
To display a generated image within ARC, you can use the Camera Device robot skill to show images. Here's a step-by-step guide on how to do this:
Add the Camera Device:
Configure the Camera Device:
Display the Image:
Using ControlCommand:
ControlCommand()function to interact with the Camera Device and other robot skills.Script Example:
Explore Further:
Feel free to ask if you need more detailed instructions or have any other questions!
@Athena It seems to me that the OpenAI Dal-e Image Skill needs a variable to store the URL that displays the generated image. Then this could be passed to the Chromium Skill and displayed.
You're correct! The OpenAI DALL-E Image Skill generates an image and provides a URL for the image. You can store this URL in a variable and then use the Chromium Skill to display the image in a web browser. Here's how you can do it:
Generate the Image:
ControlCommand()to generate an image with the OpenAI DALL-E Image Skill.Store the URL:
setVar()function to store the URL in a global variable.Display the Image with Chromium Skill:
Here's a step-by-step example:
Step 1: Generate the Image and Store the URL
Step 2: Display the Image with Chromium Skill
Additional Notes:
By following these steps, you can automate the process of generating an image and displaying it using ARC's robot skills. If you have any more questions or need further assistance, feel free to ask!
The OpenAI Dal-e Image Skill needs a variable such as API_RESPONSE_URL to allow Chromium to display the generated image. Not worth worrying about though. Works OK as it.