index_new5.html
../../../zaker_core/zaker_tpl_static/wap/tpl_guoji1.html
![]()
谷歌近期推出了名为Gemini 2.5 Flash Image的AI图像生成与编辑工具，该工具以“Nano Banana”的代号在AI领域迅速积累人气，并一度登顶LMArena排行榜。该工具已集成至Gemini AI聊天机器人和应用程序中，并迅速推动Gemini应用在App Store中名列前茅。Gemini 2.5 Flash Image不仅能够生成高质量图像，更在图像编辑方面表现出色，用户可以通过自然语言指令精确指导编辑操作，甚至可以上传多张图片进行合成，或为个人照片生成定制头像、运动卡等。该工具可通过Google AI Studio或Gemini移动应用进行访问和使用。
✨ **强大的AI图像生成与编辑能力**：Gemini 2.5 Flash Image（代号Nano Banana）是一款集图像生成与编辑于一体的AI工具。它能够根据用户的自然语言指令，精确地对图像进行编辑，甚至可以实现多图合成、为照片添加新元素（如生成头像、运动卡等）等高级功能，极大地扩展了用户在图像创作方面的可能性。
🚀 **易于访问和使用**：用户可以通过两种便捷的方式体验Gemini 2.5 Flash Image。在电脑端，可以通过免费的Google AI Studio访问，只需登录Google账号即可开始创作；在移动端，下载Google Gemini应用，在聊天界面中寻找带有香蕉图标的“Create Image”选项，即可进行图像生成和编辑。
💡 **提升创作效果的提示词建议**：为了获得最佳的图像生成和编辑效果，谷歌建议用户提供具体且描述性的提示词，而非简单的关键词列表。详细描述场景、使用与摄影相关的术语（如相机角度、镜头类型、光线效果等）能够帮助AI更准确地理解用户意图，生成更连贯、更符合预期的图像。
💰 **免费试用与付费升级**：Gemini 2.5 Flash Image提供免费使用额度，但为测试目的有限制。若要解锁全部功能或无限制使用，用户可能需要注册Google AI Pro会员，目前也提供免费试用期。这使得用户可以在付费前充分体验其强大功能。
Google launched the Nano Banana image generator in late August, and it's been building momentum through word of mouth ever since. The new model, officially dubbed Gemini 2.5 Flash Image, actually shot to the top of the LMArena AI leaderboard under its Nano Banana code name before Google officially rolled it out in its Gemini AI chatbot and app.
Users have been so excited to try Nano Banana that Gemini now tops the App Store charts, and it’s not far behind on Google Play. Why is there so much interest in this generative AI image model?
Nano Banana is a solid image maker (though not the best AI image generator, according to our tests); however, it really shines as an AI image editor. Using natural language commands, you can tell Gemini exactly how to edit your image. This isn’t exactly new, as Gemini has been able to do this for months, but Nano Banana adds additional functionality. For instance, you can upload multiple images and have Gemini combine them into a single image. You can also upload a photo of yourself, and using Nano Banana, have Gemini generate headshots, sports cards, or add new details to your original photo.
Left: Google provided this example when announcing Nano Banana. Credit: Google
Right: Credit: Google
Left: Our tech editor provided these before and after photos after giving Nano Banana a shot. Credit: Timothy Beck Werth / Mashable
Right: Credit: Timothy Beck Werth / Mashable
In any case, it's an impressive image editor (and generator), and here’s how you can use it.
How to use Google Nano Banana
The easiest way for people on a computer to play with Nano Banana is to head to Google AI Studio.
Google AI Studio is the company's free playground for users who want to try its latest AI tools. (It's a great way to try AI tools before Google locks them behind a paywall.) To get started, simply log into (or create) your Google account and then opt in to Google AI Studio. If you're already signed up for Google AI Studio, you can go straight in and start uploading, generating, and editing images.
On mobile, it’s similarly simple. Download the Google Gemini app on Android or iOS. Once you've logged in, open the app and start a new chat. Look for the option to Create Image with a banana icon. Tap that and start entering prompts or upload an image of your own.
Please note: If you're having trouble accessing Nano Banana, then you may need to sign in to your Google account and/or sign up for Google AI Studio first.
Those are the two easiest methods if you want to try Nano Banana right now. Per Google, the free version has a limited number of uses “for testing purposes” before you have to sign up for a Google AI Pro membership (a free trial is available), so make your usage count. You can also access Nano Banana in the browser version of Gemini or test it against other models at LMAerna
How to generate an image with Nano Banana
Generating an image in Nano Banana is the same as generating an image on any other image generator. You type in your request and allow the AI to do its thing. Here's how to get started:
Google AI Studio
Go to Google AI Studio (or Gemini).
Select the Try Nano Banana option.
In the given text box, type in your prompt.
Once done, hit Run.
After some time, Nano Banana will generate the image you requested.
Enter additional prompts to edit or alter the image.
Google Gemini app
Look for the banana icon in the Gemini app. Credit: Timothy Beck Werth / Mashable
Open the Google Gemini app
On the first screen that loads, you should see the option to Create Image with a banana icon. Tap it or hit the + icon and upload an image of your own.
In the text box, type your prompt.
Once done, hit the send button.
Again, Nano Banana will take some time and then generate the image you wanted.
In testing, images tended to generate in about 10 to 15 seconds, depending on the prompt. If you're not getting the type of results you imagined, try altering your prompt.
How to edit images with Nano Banana
This is pretty much the same process, but with one extra step where you upload the media that you want Nano Banana to edit.
Google AI Studio
Load into Nano Banana using the above instructions.
This time, tap the plus button on the right side of the text box.
You can choose where to source your media from. Your options are Google Drive, uploading a file, taking a photo with your webcam, and using a piece of sample media.
Once you select your image, it will upload to Nano Banana.
From here, you can ask Nano Banana to edit or alter the image how you choose using the text box to type in your prompts.
Google Gemini app
Again, follow the above steps to open a chat to create images.
Hit the plus button in the bottom left corner.
You can choose where to source your media. Your options include your camera (taking a picture), the gallery, the file browser, and Google Drive.
Once you select your image(s), they will upload in the chat window.
Type what you want Nano Banana to do with your image(s) and the AI will do its best to comply.
During testing, I found this process to be hit or miss. You have to be very specific about what you want it to do if you want to combine two images into one.
How to get the best results with Nano Banana: Be specific
For the most part, Nano Banana worked quite well. However, it does have its limitations. For example, I had it transpose my hand holding BBQ spice rub into a picture of a backyard bonfire. It did so with beautiful results. However, I asked it to darken my hand to match the lighting, and Nano Banana simply didn’t do it, showing me the same image again. So, if you give it a try, make sure to make your prompts as specific as possible to avoid mixups, and it’s otherwise a lot of fun to play with.
Google also has documentation for making the most of Nano Banana.
To get the best results, Google also recommends being specific and descriptive: "Describe the scene, don't just list keywords. The model's core strength is its deep language understanding. A narrative, descriptive paragraph will almost always produce a better, more coherent image than a list of disconnected words."
For instance, if you're hoping to generate photorealistic images, then be sure to use lots of terms related to photography, such as specific camera angles, lens or film types, and lighting ideas.
