Google ships Gemini 3.1 Flash Image and Gemini 3 Pro Image
The two new image-generation models split the market by use case: Flash Image for high-volume, cost-sensitive generation, and Pro Image for quality-first work.
NFNiko Ferrar
Published 6 Jul 2026, 07:32Updated 29 Jul 2026, 03:341 min read
While Gemini 3.5 Pro remains in development, Google released two new image-generation models on June 30: Gemini 3.1 Flash Image and Gemini 3 Pro Image.
Two tiers, two jobs
- ▸Flash Image targets high-volume, cost-sensitive generation workflows — the kind that runs thousands of times a day inside an application
- ▸Pro Image targets quality-first generation, for work where output fidelity matters more than throughput or cost
Why split the product this way
Image generation demand splits cleanly into "as many as possible, cheaply" and "as good as possible, occasionally." Shipping two dedicated models instead of one general-purpose tier lets Google price and optimize each workload separately rather than compromising on both.
Source: Google DeepMind
