Google announces that its Gemini 2.0 Flash image generation function is now open in the preview phase and that developers can create and edit images by Google AI Studio and Vertex AI using models Gemini-2.0-flash-preview-image-development. The Gemini 2.0 Flash image generation function was upgraded at the same time
#Gemini 2.0 Flash image generation upgrade content:
-
** Higher visual quality** (compared to previous experimental version)
-
** More precise text rendering**
-
** Fewer content filter interceptions** (more accessible)
** Some cases**
- Product charting background**
For example, you want to show a glass of water, and you can say, “Put this cup on the desk of the office,” and AI automatically produces a proper picture.
✍️ 2. 对话式图片修改 And you can say to one picture, “Change the grass on the lower right corner to the beach,” and AI just changes that part, and it doesn’t change anywhere else.
👩🎨 3. 实时协作绘图 It can be “co-painted” with Gemini, and it works as a team of designers (Co-Drawing Sample App).
- Text combination: Generate a diagram with text
For example, product maps with new product names and price labels can be generated for easy use in electric power suppliers and advertising.
💡 5. 帮你头脑风暴 And you said, “Give me a hand-drawn map of Markalon, and it’ll draw out the ideas in your head.
Developer achievable image generation function:
** Reconstructed: Placing products in different environments. ** Real-time co-editing of images: for example, with Co-Drawing App and Gemini. ** Local image editing: Only one part of the image is modified and the rest remains unchanged. ** Dynamic production product SKU image: new product image in combination with text. ** Creative aids: In collaboration with Gemini, design product images. ** Example code (call Gemini image generation): Developer can write this using Google’s API (Python code): ♪ From google import ♪ From google. Cliet = genai. Cliet (api_key = “Your API Key”) = client.models.generate_content(s) Model = “gemini-2.0-flash-preview-image-development” “Shows how to make a graphic lesson for Markalon.” Config=types.GenerateContentConfig( [“TEXT”, “IMAGE”] This is the first time I’ve ever seen you. I’m not sure what I’m talking about. You’ll get an AI-generated map with text, like a short teaching card.