ChatGPT Introduces Free Image Generation
- ChatGPT now offers free image creation from text prompts using the DALL-E 3 model.
- Free users can create up to two images per day, the same feature previously only available to paid users.
- The free version has some limitations, including no options to fine-tune or adjust images like other tools.
ChatGPT now lets free users generate a maximum of two images per day.
Now, ChatGPT’s free users can generate images using text prompts, thanks to the addition of the DALL-E 3 model. This feature was initially available only to ChatGPT’s Plus and Enterprise members since October 2023, but it’s now accessible to everyone, though with a catch—you can only create two images per day.
The two-image limit seems to be a way to encourage users to upgrade to the premium version. Additionally, the platform doesn’t offer any direct controls to tweak or optimize the images, something you might find on other image generators like MidJourney. After using up your two daily tries, ChatGPT will prompt you to come back after 24 hours.
Despite these limitations, the images generated are powered by the same DALL-E 3 AI model used by ChatGPT’s paid subscribers, so the quality should be quite similar. However, the final image quality still heavily depends on how well you craft your prompts. This move by OpenAI might also be an attempt to gather more data for improving DALL-E before they launch its next version.
While ChatGPT restricts image generation to two attempts per day, Microsoft’s Copilot in Bing (formerly Bing Chat) and the Microsoft Designer’s Image Creator offer similar image-generating features without any daily limits. Both Microsoft tools also use the DALL-E 3 model, which might make some users consider switching platforms unless they’re attached to other ChatGPT features, like custom GPTs, or are using the app on their phone.
On the other hand, chatbots from other tech giants, like Google’s Gemini, allow users to generate unlimited images from text prompts for free. Plus, Gemini offers at least four variations of the image per prompt, giving it a bit of an edge over ChatGPT in this area.
