Google is racing to make AI imagery faster and cheaper, even if that means accepting slightly rougher results compared to its premium models.

On June 30, Google announced Nano Banana 2 Lite, describing it as its “fastest and cheapest” AI image generator in the Nano Banana lineup.¹ The model can produce images in about four seconds and costs “under four cents per thousand images,” according to coverage of the launch.¹ It immediately became available through Google AI Studio, the Gemini API, and the Gemini Enterprise Agent Platform, replacing the original Nano Banana as a “legacy model.”¹

Shortly after the technical reveal, Google executives amplified the message on social media. CEO Sundar Pichai retweeted a post highlighting that Nano Banana 2 Lite is “extremely fast (<4s image) & cheap ($0.034 / 1K image)” and pairing it with Gemini Omni Flash, a video model positioned as “SOTA at video editing at $0.10 / sec.”Sundar Pichai (sundarpichai)² The tweet framed both as new “generative media models in the Gemini API and AI Studio,” underscoring Google’s push for a unified developer pipeline.³

Reporters quickly contextualized the launch within the broader AI image market. The Next Web described Nano Banana 2 Lite as “the fastest and cheapest model in its Nano Banana family,” aimed squarely at “rapid ideation and high-velocity developer pipelines” where latency and cost outweigh fine detail.¹ It emphasized that higher-fidelity work should still rely on Nano Banana 2 or Nano Banana Pro.¹

Ars Technica stressed the trade-offs: while Google DeepMind claims Nano Banana 2 Lite offers “the best balance of quality and speed,” the images “may not look as good” as those from heavier models and can struggle with small text, infographics, and character consistency.⁴ Even so, its roughly four‑second turnaround and $0.034 per 1,000 images mean developers “will pay a lot less,” making it attractive for rapid prototyping and design exploration.⁴