Gemini 2.0 Flash (native image generation)
2025-03-12 · Publisher : Google · Domain : Artificial Intelligence
Google opens Gemini 2.0 Flash's native image generation and editing to developers in an experimental version, with text and images interleaved in a single response.
What changes
- Native image output in the model: no more call to a separate generation model.
- Conversational image editing over several turns of dialogue.
- Illustrated stories with characters and settings that stay consistent from one image to the next.
- World knowledge and reasoning put to use to produce accurate images (illustrating a recipe, for example).
- Text rendering in images superior to competing models according to Google's internal benchmarks.
### Specifications - Availability: Google AI Studio and the Gemini API, in all regions supported by AI Studio. - API identifier at launch: `gemini-2.0-flash-exp`. A preview `gemini-2.0-flash-preview-image-generation` followed on May 7, 2025; both image models (`-exp-image-generation` and `-preview-image-generation`) were discontinued on November 14, 2025. - First preview reserved for trusted testers in December 2024. ### Pricing Pricing not disclosed at launch (experimental). ### Benchmarks Internal comparisons cited without figures in the post. ### Key takeaways A preparatory step toward Nano Banana (Gemini 2.5 Flash Image): image becomes an output modality of the language model, which simplifies editing and illustration use cases.
Links
Tags : Image, LLM, Google, Gemini, Flash, Expérimental