Hey everyone!

This week’s GenAI Guide focuses on exploring Gemini’s native image editing— a feature that lets you customize your photos and images through simple back and forth conversations with the model.

As someone who’s constantly working with unique visuals yet lacking any meaningful artistic talent, I’m always looking for tools that help bridge the gap. What used to be a difficult and time consuming task has become something I always look forward to. But one aspect of AI image generation has consistently lagged behind: image editing.

Luckily Google’s latest update to the Gemini app has brought native image generation to everyone. This update allows you to edit images by simply chatting with the model. 

Getting Started with Native Image Editing

Accessing Google’s new native image generation is simple and free. Just head to the Gemini app and sign into your Google account. From there you can start editing. It's that easy!

Currently Gemini supports editing of both uploaded images and those created within the app. The following image types are supported on Gemini:

• PNG 

• JPEG 

• WEBP 

• HEIC 

• HEIF

Tips for Success

• Be Specific: I’ve learned that whether you’re adding a new element to an image or changing the color scheme, giving a detailed explanation of what you want will significantly speed up the process. 

• Iterative Design: This is where Gemini shines! Some of my best results have come from slowly changing an image until it's perfect. As you can see from the example below, I was able to ask Gemini to make one set of edits before having it build off of what was created. 

• Experiment: Like all AI tools, native image generation requires hands-on exploration. Don’t be afraid to try wacky prompts to see what happens. The more you experiment, the better the images you create will be. 

• Identify Limitations: While Gemini’s native image generation is impressive, it's far from perfect. I’d suggest learning to identify what the model can and cannot do. For example, I’ve noticed that removing people to an image doesn’t work well because the model struggles to distinguish between people.

• Create Workflows: While not everyone will be creating images within Gemini, you should still include it within your image workflows. Whether you're touching up the colors, adding a logo, or creating something completely new, I’d highly recommend using Gemini to create the perfect visuals. (For a list of other great AI image generators to incorporate in your workflow, check out our 2025 Image Generator Guide.)

• Create Images with Text: One of the most frustrating aspects of generating images with AI is that they can’t include text. The latest Gemini updates solve this problem by allowing you to easily include text within any images. 

Pros and Cons 

Like any tool, Gemini's native image editing has its own strengths and weaknesses. I’ll try to cover everything that I’ve noticed while working with this tool.

Pros : 

• Uses natural language. Instead of needing a complex editing tool, your images can be refined through a conversation with the model. This is great because it can use your prior conversation to help shape the image. 

• Adding Text. Being able to add and edit text is such an important feature. Many other AI models struggle to generate actual words (let alone sentences). This is my favorite aspect of this tool.

• Accessible to everyone. Anyone with a Google account can access this tool for free. I’d recommend trying this tool out and seeing if it meets your expectations. 

• Creative. I’ve found that the model can be incredibly creative if you let it. Some of my best results came from me allowing the model to be creative. Check out this reddit post that showcases some examples of Gemini in action!

• High rate limits. Another significant benefit of this tool is that you can create so many images per day. While Google doesn’t publish their rate limits, I’ve been able to create more than a dozen images a day. 

Cons:

• Current Limitations. While this first iteration is impressive, the tool is still lacking polish. Often I’ve had to repeat a request multiple times to get the result I’m looking for which can be frustrating.

• Bugs. Unfortunately I’ve run into a number of bugs and glitches while using this tool. Anything from Gemini just randomly combining past images together to telling me it won’t create an image of a car because it's against TOS. 

• Inconsistent. Overall my biggest issue with Gemini’s native image editing is the inconsistency. Sometimes it creates incredible visuals on the first try while other times it can’t make small tweaks. While most AI image generators are like this, I found that Gemini would swing between great and bad more often than most. 

Conclusion

I’m always excited to discover new AI tools that help make people’s lives easier or gives them the ability to do things they otherwise couldn’t. Through this lens, Gemini’s image editing is remarkable. It's another step toward bringing the power of custom image creation to everyone. I love the idea that people without an artistic background (I include myself in that group) can now create some super cool visuals! Whether I’m using AI to create images for articles or just letting my imagination run wild, the ability to customize my creations though natural language has helped me shape my work. 

So, what are you waiting for? Fire up Gemini, unleash your inner artist, and start creating!