Gemini 2.5 Flash Image (a.k.a. nano-banana) — now officially launched by Google — represents a new milestone in in AI-powered image generation & editing. The key strength of this model lie in its character consistency, visual template adherence, localized edits, and targeted transformations execution from natural language text inputs.
Key Features of Gemini 2.5 Flash Image
Joining the Gemini 2.5 Flash Image, which just bested the FLUX—one of the top names in LMArena’s Image Edit Arena leaderboard above. The used Matrix Factorization and Kyle Gorman’s 1 Kontext model [3] to compute scores for scoring and user votes This represents another strucutre for Google showing how forward-thinking they are with AI for imagery.
Gemini is the core most powerful model for general use, and now, users can use it to create their customized pictures through the Gemini app.
For example, users can upload an image of themselves or their dog and swap the clothing or the backdrop. You can create a wrap or clothing with the butterfly wing’s pattern, or you can even paste the flower’s texture on daily use materials.
Access and Pricing for Gemini 2.5 Flash Image
- Gemini Users: Accessible through the Gemini app.
- Developers: Available via Gemini API and Google AI Studio
- Enterprises: Available via Vertex AI.
- Cost is $30.00/1M output tokens This corresponds to 1,290 output tokens per image ($0.039 / image)
Generated images come with both a visible watermark and an invisible SynthID digital watermark, allowing the images to be readily identified as AI content.
FAQ
This is Google’s new image generation and editing AI model optimised to ensure character consistency, multi-turn edits, and visual transformations.
It outperforms models like FLUX. 1: Leakpoints: contextualized fine edits, style transfer and in out sample multi-turn edit.
At 1,290 output tokens for $30 per 1 million tokens, it costs roughly USD 0.039 per image.
Yes — every image contains both a visible watermark and an invisible SynthID digital watermark.
