ChatGPT has been incredibly helpful in various areas, assisting us greatly with tasks such as summarizing articles and organizing or composing emails. What about using ChatGPT with images? Here, we provide a comprehensive guide on using ChatGPT with image inputs.
Does ChatGPT accept images as input?
Yes, according to OpenAI’s announcement made on September 25, 2023, ChatGPT Plus and ChatGPT Enterprise users will now be able to send image inputs to ChatGPT through the GPT-4 model.
Previously, ChatGPT only accepted text input. The sole exception was the Code Interpreter model, which allowed you to upload a file and execute Python code.
On March 14, 2023, OpenAI released the GPT-4 model. The most significant feature that sets the GPT-4 model apart from its predecessors is its multimodal capability. This means it can accept input in both text and image formats. Thanks to this feature of GPT-4, you will now be able to use the image upload capability in ChatGPT as well.
Also read: GPT-4V(ision): How to give GPT-4 image input?
Note that ChatGPT and GPT-4 are not the same thing. Even if you could call the GPT-4 model through ChatGPT, these are two different concepts.
Also read: How to call GPT-4 model in ChatGPT?
How to upload a image in ChatGPT?
Only ChatGPT Plus members can upload images to ChatGPT using the GPT-4 model. You can use this feature on both mobile apps and web browsers.
First, select the GPT-4 model.
Click on the image icon in the bottom left corner of the message box. Select the photo you want to upload.
You can choose multiple images. After selecting and uploading your choices, type your message in the message box and send.
ChatGPT with Images
Now, you have the capability to share one or more images with ChatGPT. For instance, if your bicycle isn’t working, you can take a photo of it and send it to ChatGPT to troubleshoot the issue. Or, you can snap a picture of the inside of your fridge and inquire about meal options with the available ingredients.
With the image upload feature, you can now upload images to ChatGPT, enabling you to have it analyze or answer questions related to the uploaded images.
How to send image inputs to ChatGPT?
ChatGPT Plus users can use the image upload feature both in their web browsers and through the mobile application.
Also read: 5+ Exciting Examples of Using ChatGPT with Images
To upload images to ChatGPT, follow these steps after logging into your account:
- Tap the “Photo” button in the text box on the screen and select “Capture or Choose an Image.”
- If you’re using the ChatGPT iOS or Android mobile app, tap the plus button first.
The images you upload can be photographs, screenshots, or files containing both images and text.
You also have the option to upload multiple photos and use the drawing tool to make annotations or highlight specific areas on the uploaded image.
How to focus a specific part of the image in ChatGPT?
To focus on a specific part of the image, you can use the drawing tool in the mobile app.
After uploading an image, you can use the drawing tool to make annotations on the image and focus on the specific areas you want to inquire about when interacting with ChatGPT.
The areas enclosed within circles or marked with arrows in an image will be considered points of focus by ChatGPT.
Is it only ChatGPT Plus members who can use the image upload feature?
Yes, currently, only ChatGPT Plus and ChatGPT Enterprise paid subscriptions can use the image upload feature. If you are using ChatGPT for free, you cannot use the image input feature at the moment.
ChatGPT vs GPT-4: What is the difference?
GPT-4 is actually the latest model in the series of OpenAI’s GPT models, and it can be called via API. On the other hand, ChatGPT is an application specialized in conversational dialogue, which uses the GPT-3.5 model as its base model.
ChatGPT has been trained specifically for interactive conversations. It allows you to access the GPT-4 model through a user-friendly interface without needing to directly interact with the GPT-4 API.
How to send input files to ChatGPT?
With the Code Interpreter model exclusive to ChatGPT Plus members, you can upload files to ChatGPT.
You can request ChatGPT to analyze the data in the file, execute Python code provided in the file, and perform analyses based on the data in the file. To use this model, it is necessary to use GPT-4 model in ChatGPT.
Uploading files indeed provides significant advantages because ChatGPT has a character limit, and inputs exceeding that limit may not be understood or processed properly. By uploading files, you can overcome this limitation and request ChatGPT to handle and analyze data with high character counts.
Does ChatGPT generate images?
The answer is both yes and no. ChatGPT, in the traditional sense, cannot generate images. To create images, you would need to use other models specialized in that task. One example is OpenAI’s DALL-E 3 model, designed specifically for image generation.
However, through ChatGPT Plugins, you can now access third-party applications. One of these is the DALL-E-3 Plugin. With this feature, you can have ChatGPT generate images.
Learn More: How to use ChatGPT Plugins?
ChatGPT can also generate images in SVG (Scalable Vector Graphics) and ASCII formats. SVG format uses mathematical expressions for vector-based drawings, while ASCII format represents images using characters, allowing ChatGPT to handle these formats.
Learn more: How to create images in ASCII and SVG format in ChatGPT?
Another tool that can assist with image generation is Bing Chat, which also uses the GPT-4 model by OpenAI. By combining Bing Chat with ChatGPT, you can even create unique logos with custom features, almost like a graphic designer.