Google’s Nano Banana is live for developers through the Gemini API and Google AI Studio in preview. It is an image model officially called Gemini 2.5 Flash Image, and you can try it for free with rate limits. There is also a special 48-hour free unlock running this weekend for builders.
First, what Nano Banana actually is
Nano Banana is a nickname for Gemini 2.5 Flash Image. It generates and edits images with strong character consistency, multi-image fusion and quick local edits. It ships as the gemini-2.5-flash-image-preview model in the API and AI Studio. All outputs have SynthID watermarking.
Free trial. What you get today
About video. Quick clarity so you do not get stuck
Nano Banana itself is for images. For image-to-video in the Gemini stack you use Veo 3 via the Gemini API. Google’s own docs show image to video recipes with Veo. Media reports also confirm the photo to video feature in Gemini is powered by Veo 3. So your workflow is simple. Make images with NanoBanana, then pass them to Veo for video.
Try it now. Two fast paths
- Google AI Studio
- Open the model page, pick gemini-2.5-flash-image-preview, type a prompt, optionally upload a base photo, and generate.
- Gemini API
- Set
model="gemini-2.5-flash-image-preview"and send your prompt and image. This endpoint returns images. For video generation switch to the Veo video endpoint as shown in the docs.
- Set
Key details at a glance
- Model name gemini-2.5-flash-image-preview.
- Type image generation and editing. Not a native video model.
- Free access API free tier and current 48-hour unlock for hackathon. Google AI for DevelopersGitHub
- Image-to-video use Veo 3 in the Gemini API.
