Google announced a significant upgrade for Gemini, its in-house artificial intelligence (AI) model, on Wednesday. The company announced that the image generation capability of the chatbot will now be handled by the Imagen 3 AI model for all users. Imagen 3 is the Mountain View-based tech giant’s latest and most capable image generation model. Apart from the Gemini app, the feature is also being extended to the API version of Gemini to let developers build apps and experiences based on this capability.

Gemini Users Get Access to Imagen 3 AI Model

In a post on X (formerly known as Twitter), the official handle of the Google Gemini App revealed that all users, including those on the free tier, will be able to generate images using Imagen 3. The post highlighted that the AI model offers a high degree of photorealism, better prompt adherence, and adds fewer unwanted elements to images.

Gadgets 360 staff members were able to verify that the Gemini app is indeed using Imagen 3 to generate images. To test its capabilities and compare it with Meta AI, we gave both chatbots the same prompt. The prompt was, “Draw an image of a golden retriever dog sitting on a train berth, looking out through the window at the Alps. The train has a wooden interior and the seats are green in colour. All other passengers on the train are also animals. One human conductor is checking for tickets.”

Meta AI vs Gemini

 

The generated images can be seen above. While both AI models failed to incorporate one or more elements instructed in the prompt, Gemini was able to incorporate more elements. Additionally, while Meta AI generates images in 1280 x 1280 resolution, Imagen 3 images are generated in 2048 x 2048 resolution.

Imagen 3 can generate images in a wide range of styles such as photorealistic, textured oil paintings, and claymation scenes. Users can also request images to appear as if it has been taken from a specific camera such as a Nikon DSLR, GoPro style, wide-angle lens, and more.

Google has said that the AI model comes with inbuilt safeguards to reduce the risk of deepfakes. Every generated image also comes watermarked with SynthID, a technology that adds an invisible AI label within the pixels of the image. It cannot be cropped out or removed and is present even in screenshots.

Google announced a significant upgrade for Gemini, its in-house artificial intelligence (AI) model, on Wednesday. The company announced that the image generation capability of the chatbot will now be handled by the Imagen 3 AI model for all users. Imagen 3 is the Mountain View-based tech giant’s latest and most capable image generation model. Apart from the Gemini app, the feature is also being extended to the API version of Gemini to let developers build apps and experiences based on this capability.

Gemini Users Get Access to Imagen 3 AI Model

In a post on X (formerly known as Twitter), the official handle of the Google Gemini App revealed that all users, including those on the free tier, will be able to generate images using Imagen 3. The post highlighted that the AI model offers a high degree of photorealism, better prompt adherence, and adds fewer unwanted elements to images.

Gadgets 360 staff members were able to verify that the Gemini app is indeed using Imagen 3 to generate images. To test its capabilities and compare it with Meta AI, we gave both chatbots the same prompt. The prompt was, “Draw an image of a golden retriever dog sitting on a train berth, looking out through the window at the Alps. The train has a wooden interior and the seats are green in colour. All other passengers on the train are also animals. One human conductor is checking for tickets.”

Meta AI vs Gemini

 

The generated images can be seen above. While both AI models failed to incorporate one or more elements instructed in the prompt, Gemini was able to incorporate more elements. Additionally, while Meta AI generates images in 1280 x 1280 resolution, Imagen 3 images are generated in 2048 x 2048 resolution.

Imagen 3 can generate images in a wide range of styles such as photorealistic, textured oil paintings, and claymation scenes. Users can also request images to appear as if it has been taken from a specific camera such as a Nikon DSLR, GoPro style, wide-angle lens, and more.

Google has said that the AI model comes with inbuilt safeguards to reduce the risk of deepfakes. Every generated image also comes watermarked with SynthID, a technology that adds an invisible AI label within the pixels of the image. It cannot be cropped out or removed and is present even in screenshots.

Google is reportedly working on improving the image generation capabilities of the Gemini chatbot. As per a new report, the Mountain View-based tech giant is adding an inline image editing feature to the artificial intelligence (AI) bot. The feature was spotted in the latest beta version of the Google app, and while it is reportedly being developed for smartphones, it is unclear whether it will also be available on the web client. Notably, the Microsoft Designer platform already offers an inline editing feature, which is powered by Copilot.

Gemini Inline Image Editing

According to an Android Authority report, the new feature was spotted in the Google for Android beta app version 15.40.31.29. It was seen by the publication during an Android application packaging (APK) teardown process. Those who have signed up for the Google beta programme will not be able to test it out as it’s not a visible feature.

So far, editing an image using Gemini has been a tough task. This is because if a user did not like a particular detail about the image, they would have to add more details in a follow-up prompt to generate another iteration. However, the next iteration could remove the good parts from the earlier image and introduce new aberrations as well. There was no way to retain the good parts while eliminating the not-so-good ones.

But, this might change with the new feature. Based on a demonstration video shared by the publication, the feature works based on a combination of two interactions by the user. Once an image has been generated, the user can highlight a portion of the image and add a follow-up prompt, and Gemini will only change that portion of the image, leaving the rest of it untouched.

However, as users can see in the video demo, the feature is far from completion. In several instances, it still changed the whole image despite being instructed to only change a part of it. But it is possible that when the feature is rolled out to users, these issues could be taken care of.

Google is reportedly working on improving the image generation capabilities of the Gemini chatbot. As per a new report, the Mountain View-based tech giant is adding an inline image editing feature to the artificial intelligence (AI) bot. The feature was spotted in the latest beta version of the Google app, and while it is reportedly being developed for smartphones, it is unclear whether it will also be available on the web client. Notably, the Microsoft Designer platform already offers an inline editing feature, which is powered by Copilot.

Gemini Inline Image Editing

According to an Android Authority report, the new feature was spotted in the Google for Android beta app version 15.40.31.29. It was seen by the publication during an Android application packaging (APK) teardown process. Those who have signed up for the Google beta programme will not be able to test it out as it’s not a visible feature.

So far, editing an image using Gemini has been a tough task. This is because if a user did not like a particular detail about the image, they would have to add more details in a follow-up prompt to generate another iteration. However, the next iteration could remove the good parts from the earlier image and introduce new aberrations as well. There was no way to retain the good parts while eliminating the not-so-good ones.

But, this might change with the new feature. Based on a demonstration video shared by the publication, the feature works based on a combination of two interactions by the user. Once an image has been generated, the user can highlight a portion of the image and add a follow-up prompt, and Gemini will only change that portion of the image, leaving the rest of it untouched.

However, as users can see in the video demo, the feature is far from completion. In several instances, it still changed the whole image despite being instructed to only change a part of it. But it is possible that when the feature is rolled out to users, these issues could be taken care of.

Meta AI features were a major highlight during the company’s annual Meta Connect event on Wednesday. The company introduced the new voice chat feature and added new artificial intelligence features to the Ray-Ban Meta smart glasses. The social media giant also expanded the experimental ‘Imagine’ feature to Facebook, Instagram, and Messenger. The feature was unveiled earlier this year, allowing users to recreate their profile picture in different AI-generated styles. This is now being rolled out to all users. Additionally, Meta is also testing a feature that will auto-generate images and recommend users to share them to their feed.

In a newsroom post, the social media giant shared details about how this feature would work. So far, users can use Meta AI to generate images from text prompts. Users can also share an image as an input and ask the AI to generate images based on that. However, the company did not let users use their images of their profile pictures to generate images.

This is now changing as at the Meta Connect 2024, the company announced that the ‘Imagine’ feature will be available to all Facebook and Instagram users and this feature will allow users to create posts as well as Stories.

Put simply, users can now ask Meta AI to ‘Imagine’ them as a superhero, a royal monarch, or an astronaut, and the AI will do so. Based on the images shared by the company, it appears the AI-generated images will not be very realistic and instead appear digitally created. Meta is also adding AI-generated watermarks on these images as well.

In Instagram DM and Messenger, users will be now able to use AI to generate personalised chat themes. Users could earlier change the background colour and chat bubble, but this will offer a higher level of customisation.

Additionally, Meta is also testing a recommendation feature involving the Imagine feature. The company plans to show images auto-generated AI images of the user’s profile picture in the feed of Instagram and Facebook. The generated images will be based on the user’s interests or current trends. If the user likes the image, they can share it on their feed or Stories. They can also tap on the image to make changes to the suggested image.