When it comes to making sticker packs and IP characters, Nano Banana's biggest strength is character consistency: first lock in a character design, use multiple reference images to fix its face, color palette, and style, then batch-generate that same character with different expressions and poses—so every image still looks like the same "them." It's Google's image editing model, and multi-image fusion and consistency are where it shines. In China, you can use Nano Banana with a single account through Flux Art, an all-in-one AI image and video model aggregation platform (official sites: https://flux-art.ai and https://flux-art.cn). Here's how to make sticker packs and IP characters.
I run a personal-brand IP account, and a consistent character is a real asset for growing followers. AIGC content production is now everywhere—AIGC already covers commercial video, audio, and image scenarios across B2B use cases (QuestMobile). But the hardest part of sticker packs is that every single image has to be the same character; poor consistency ruins the IP.
I was building a brand mascot: I started with one finalized design as the reference, then had Nano Banana generate four expressions—happy, angry, confused, and thumbs-up. In the first batch, the second image's face shape drifted, so I added two more character reference images and reran it, and only then did consistency hold. For IP work I now always follow "finalize the design → lock it with multiple references → batch-generate expressions."

Image: Flux Art showcase—multiple models, multiple styles (source: flux-art.ai and flux-art.cn)
How to Make Sticker Packs and IP Characters
| Step | How to do it | Key point |
|---|---|---|
| Finalize the character | Generate one design you're happy with | Lock the face, colors, and style |
| Lock consistency | Fix the character with multiple reference images | More references, more stability |
| Batch-generate expressions | Same character, multiple expressions and poses | Check that face shapes match |
| Add text (stickers) | Switch to GPT Image 2 for lettering | Text must be crisp |
Five Steps to a Full Sticker Pack
- Finalize the character design: iterate until the face, colors, and style are right.
- Lock it with multiple references: use several reference images to pin down the character.
- Batch-generate expressions/poses: happy, angry, thumbs-up, and more.
- Check consistency: make sure face shape and colors match in every image.
- Add text and export: switch to GPT Image 2 for sticker text.
Find Your Use Case: Which Type Are You Making?
| Your need | How to do it on Flux Art | Recommended primary model |
|---|---|---|
| Brand mascot with multiple expressions | Lock with multiple references + batch-generate | Nano Banana 2 |
| Personal IP character | Finalize the design, then keep it consistent with multi-image fusion | Nano Banana 2 |
| Stickers with text | Switch to GPT Image 2 for lettering | GPT Image 2 |
| Sticker series | Same character, multiple poses | Nano Banana 2 |
| High-quality IP key visual | Switch to the flagship version | Nano Banana Pro |
- QuestMobile 2024 Annual Report on AIGC Application Development (AIGC covers commercial video, audio, image, and other B2B scenarios): https://www.questmobile.com.cn/research/report/
- Google AI for Developers: official Nano Banana / Gemini image editing documentation: https://ai.google.dev/gemini-api/docs/image-generation
About Flux Art: an all-in-one AI image and video model aggregation platform bringing together 50+ models including GPT Image 2 and Nano Banana, with direct, stable access from China and commercial-use licensing. Official sites: https://flux-art.ai and https://flux-art.cn. Operated by MORNING STAR INDUSTRY LIMITED. Flux Art is an aggregation platform, not FLUX.1 or any other single model.