Google made its move on June 30 with two new generative models: Nano Banana 2 Lite, its fastest and cheapest image generator, and Gemini Omni Flash, a video model you edit by talking to it. The first competes on price; the second aims at something more interesting: making video editing a conversation. Here’s what’s real, verified, and our take.
Nano Banana 2 Lite: images in 4 seconds
- The fastest and cheapest of the series: it generates an image in about 4 seconds for around $0.034 (1K resolution), according to Google’s announcement.
- Technical name: Gemini 3.1 Flash Lite Image — the light variant of the family you already knew as Nano Banana.
- What it’s genuinely for: volume. Thumbnails, variations, A/B tests, drafts. Where cost used to make you think twice, now you iterate without watching the meter.
Gemini Omni Flash: video you edit by talking
- Generates and edits video from text, image or video, with native audio on every output and character and style consistency.
- Conversational editing: swap a character, relight a scene or change the angle by asking in natural language, while keeping the original audio and video tracks.
- Price and limits: $0.10 per second of generated video, clips up to 10 seconds, in public preview on Google AI Studio and the Gemini API.
Our take: editing is the news, not generation
- Generating video no longer surprises anyone; editing it by talking does. “Swap the character and keep my audio” is exactly the invisible work we said would transform the industry before “AI cinema” ever does.
- The price defines the use: at $0.10/second, a 10-second clip costs $1. For serious production that adds up fast; for prototypes, b-roll and testing, it’s new territory.
- The usual warning still applies: these are preview models and the 10-second limit rules. Test YOUR use case before promising a client anything, and remember the red lines: no recreating real people without permission.
Our honest reading: Google is attacking both ends at once — cost per image downwards and editing friction downwards. The generative video battle is decided less and less by clip quality and more and more by workflow.
Frequently asked questions
Can I try them now?
Yes: both are available in Google AI Studio and via the Gemini API (Omni Flash in public preview). For casual use, wait until they reach the consumer apps.
Do they replace Veo?
Not exactly: Omni Flash prioritises speed and conversational editing; the large video models remain the maximum-quality reference. Different tools for different phases.
Conclusion
Nearly free images and video you fix by conversing: both announcements point the same way — removing friction. For full context, see the real state of generative video and what Gemini 3 and the Nano Banana family brought.
