Nano Banana 2 INSANE Product Photos
Evaluating Nano Banana 2 and Soul 2 demonstrates how modern image models handle complex typography and multi-character consistency for commercial product workflows.
What this lesson covers
Thomas Lundström stress-tests Nano Banana 2 for product photography, highlighting its enhanced text rendering and speed over previous versions. He also compares its subject consistency with Higgsfield AI's Soul 2 using custom Soul ID characters.
Key takeaways from the creator
AI-extracted notes, not independently verified product claims. Timestamp links let you check each point in the original video.
- 00:00 ↗
Nano Banana 2 introduces upgrades to visual fidelity, world knowledge, label text rendering, and instruction following, running faster than the Pro version.
- 01:15 ↗
Testing reveals that Nano Banana 2 accurately renders complex graphic and text elements on product packaging from simple iPhone reference photos.
- 06:50 ↗
Higgsfield AI's Soul 2 model offers superior character consistency compared to Nano Banana 2 by letting users create a Soul ID using up to 80 reference images.
- 07:25 ↗
Combining Soul 2's character consistency with Nano Banana's editing features allows creators to maintain uniform color palettes and consistent clothing across product shoots.
Workflow outlined in the video
- Upload a clear reference photo of your product into Nano Banana 2 to test text and label rendering accuracy.
- Utilize Higgsfield AI's Soul 2 to create a Soul ID using multiple angle photos if you need strict character consistency.
- Combine consistent character generation with image editing features to maintain color and style across multiple generated assets.
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
21,852 views captured 24 October 2026. ReelStack recommendation score: 52/100. Formula: radar-v5. This is ReelStack’s calculation, not a YouTube rating or a measure of factual accuracy.
Show the calculation
- 65% performance: 2.48× current views vs the channel's recent median; same-age history is not yet available. View-evidence factor 96% (views / (views + 1,000))
- 25% freshness: 0/100 with a 14-day half-life
- Momentum pending: collecting daily snapshots; its 20% weight goes to observed performance, not free points
- 10% engagement: 79/100; likes + 4× comments, smoothed with a 500-view neutral prior