Breaking

Google Releases Nano Banana 2 Lite Image Model and Omni Flash Video Model

June 30, 2026 at 12:25 EDT

Media Generation
Foundation Models

On June 30, 2026, Google DeepMind released the image-generation model "Nano Banana 2 Lite" and the video-generation model "Gemini Omni Flash" for the Gemini API and Google AI Studio. Both are available immediately, with the announcement emphasizing low price and low latency.

June 30, 2026 · Google DeepMind

Two New "Lite" & "Flash" Models Make AI Media Generation Cheap and Fast

Google DeepMind ships Nano Banana 2 Lite for image generation and Gemini Omni Flash for video — both available immediately, both built around low price and low latency.

5×

Faster than the prior Nano Banana 2

‑50%

Cheaper than the prior image model

<4s

To generate or edit one image

$0.034

Per 1,000 images generated

Image speed: per‑image latency

Nano Banana 2 Lite cuts time per image to roughly one‑fifth of its predecessor.

~5×

Nano Banana 2

1× (<4s)

Nano Banana 2 Lite

Text‑to‑image generation & editing
Under 4 sec per image
$0.034 per 1,000 images
Fastest, leanest of the Gemini Image lineup

Gemini Omni Flash

Video generation & conversational editing
Text‑, image‑ & reference‑to‑video
$0.10 per second (matches Veo 3.1 Fast)
First "Flash" model in the Omni family

The recommended single‑API workflow

Generate still image
Nano Banana 2 Lite

→

Animate as video
Omni Flash

→

Edit in natural language
conversational

The takeaway

At these per‑unit prices, image generation could become a default feature rather than a premium one.

Open questions remain: real‑world quality consistency and the precision of long‑form video editing — the Flash version is optimized for shorter clips, so matching models to use cases will be the next debate.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.