Google dropped two models yesterday that change the math on AI-generated content. Nano Banana 2 Lite generates images in 4 seconds for $0.034 each. Gemini Omni Flash lets you edit video by talking to it. Together, they make something that was expensive and slow feel almost instant.
What Nano Banana 2 Lite Does
Nano Banana 2 Lite is the newest model in Google’s image generation family. It does one thing really well: speed. Text-to-image in about 4 seconds. Cost per image at 1K resolution: $0.034. That is roughly half the price of Nano Banana 2 ($0.067) and a quarter of Nano Banana Pro ($0.134).
For context, that means generating 100 product shots costs $3.40. A thousand social graphics: $34. Running an A/B test with 50 ad variations: under two dollars in generation costs. The price is low enough that you can use it for things you would never bother with before, like generating entire mood boards, testing colour palettes across dozens of variations, or filling a product catalogue with AI-assisted imagery.
The tradeoff is that Nano Banana 2 Lite sits below Nano Banana 2 (the generalist) and Nano Banana Pro (the professional model) in quality. But Google says prompt adherence, character consistency, and text rendering inside images remain strong. It is built for the ideation and drafting phase, not final polished assets.
Nano Banana 2 Lite is available now in Google AI Studio, the Gemini API, and the Gemini Enterprise Agent Platform. It is also rolling out across Google’s consumer products: AI Mode in Search, the Gemini app, NotebookLM, Google Photos, Stitch, Google Flow, and Google Ads.
What Gemini Omni Flash Does
The bigger news for anyone making video content is Gemini Omni Flash. This is the model that combines Gemini’s multimodal reasoning with video generation and conversational editing. It is now available to developers via the Gemini API for the first time.
What makes Omni Flash different from other video generation tools is the editing approach. You do not just type a prompt and get a clip. You can have a conversation with it. Generate a video, then say “make the background warmer” or “slow down the camera pan” or “add text overlay that says 50% off” and it refines the output. You can mix text, images, and video as inputs and reference them in your prompts. The model draws on Gemini’s general knowledge (history, biology, narrative logic) to construct scenes that make sense.
Pricing is $0.10 per second of video output, matching Veo 3.1 Fast. A 10-second clip costs about a dollar. Currently limited to 10-second generations with more duration coming soon.
Where It Gets Interesting: Chaining Them Together
Google’s real pitch is using both models in sequence. Generate a product image with Nano Banana 2 Lite in 4 seconds, then pass it straight into Gemini Omni Flash to animate it into a video. The Interactions API preserves session context, letting you make up to three consecutive edits in a single conversation.
Google has three demo apps showing this workflow:
- Anywhere takes a selfie and places you at famous landmarks, then animates the result into a short video.
- Space Lift photographs your room, generates interior design concepts, and turns them into walkthrough videos.
- Omni Product Studio converts static product photos into e-commerce product videos.
That last one is the workflow most relevant to anyone selling products online. Upload a product photo, describe the video you want, and Omni Flash generates a clip. Then refine it conversationally until it looks right. No video editing timeline. No keyframes. Just text instructions.
What to Look Out For
The current limitations are worth knowing. Omni Flash only generates 10-second clips right now. Audio references and scene extensions are not supported yet. The API schema accepts video references up to 3 seconds, but Google says the model does not process them correctly yet. Character consistency across scene changes and camera movements is still limited.
Nano Banana 2 Lite is fast and cheap, but it is a drafting model. If you need pixel-perfect output for a client deliverable, you still want Nano Banana Pro. The Lite model is for the volume phase: generating dozens of options, rapid prototyping, filling gaps in a content calendar.
Both models use SynthID watermarks to tag AI-generated content, with verification through the Gemini app, Gemini in Chrome, or Google Search.
What This Unlocks
Two things change immediately:
1. Volume becomes viable. At $0.034 per image, you can afford to generate assets you would never have commissioned. Every product variant, every ad format, every social platform dimension. The cost barrier that made AI image generation a “special occasion” tool is gone.
2. Video iteration becomes conversational. Instead of rendering a clip, watching it, going back to the prompt, and starting over, you can now refine in real time. Say what you want changed. See the result. Say the next thing. This is how creative people actually work: they iterate. The tool finally matches the process.
For small businesses and solo creators, this is the moment where AI content production shifts from “expensive experiment” to “daily workflow.” The images are fast enough to use in live brainstorming sessions. The video editing is intuitive enough that you do not need to learn a timeline editor. You just describe what you want.
Want more like this? Join MOKU Club for free. Weekly resources, early access to new guides, and occasional templates you can actually use. Join below.
