10 Nano BANANA Prompts You Need To Try Right Now!
Learning specialized prompting techniques for emerging models like Nano Banana helps creators achieve high character consistency and execute complex multi-image blending for video pipelines.
What this lesson covers
This tutorial demonstrates ten advanced prompting workflows using the Nano Banana image model across platforms like OpenArt and HeyGen. It covers multi-image prompting, character transformation into news anchors, and integrating with video lip-sync tools.
Key takeaways from the creator
AI-extracted notes, not independently verified product claims. Timestamp links let you check each point in the original video.
- 01:43 ↗
Multi-image prompting allows users to combine a reference image of themselves with various objects, clothing, and backgrounds into a single coherent prompt output inside Nano Banana.
- 03:35 ↗
When generating widescreen 16:9 outputs, using a 16:9 input reference image avoids awkward cropping issues that occur when upscaling 9:16 screenshots.
- 05:55 ↗
Upscaling generated images two times before feeding them into video lip-sync tools like Omnihuman or Hedra significantly improves final visual fidelity.
- 06:23 ↗
Lip-sync tools offer varying duration limits per platform, ranging from 15 seconds on Omnihuman up to 90 seconds on OpenArt.
Workflow outlined in the video
- Combine your reference character and desired object images on a canvas tool like Canva or Photoshop before generation.
- Upload your reference image into Nano Banana via OpenArt and write a detailed descriptive prompt covering subject, apparel, and background.
- Set output aspect ratios explicitly to match your reference dimensions to prevent unwanted cropping.
- Upscale your final chosen image before sending it to a lip-sync tool like Hedra or Omnihuman.
Before you use this workflow
These notes describe the source video at its publication date. Model access, pricing, connectors and interfaces may have changed. Check the original source and the provider’s current documentation before installing an add-on, connecting an account or spending credits.
ReelStack has not independently tested this workflow. Preview one representative shot and check motion, continuity and output quality before applying it to a full production. No result, cost saving or model capability is guaranteed.
Explore the tools
How to interpret the numbers
51,143 views captured 23 October 2026. ReelStack recommendation score: 14/100. Formula: radar-v5. This is ReelStack’s calculation, not a YouTube rating or a measure of factual accuracy.
Show the calculation
- 45% performance: 0.31× views vs 4 other channel videos observed at a similar age. View-evidence factor 98% (views / (views + 1,000))
- 25% freshness: 0/100 with a 14-day half-life
- 20% momentum: 3 views/day vs 14 channel baseline (4 comparable uploads), with the same view-evidence factor
- 10% engagement: 4/100; likes + 4× comments, smoothed with a 500-view neutral prior