NanoBananas - Nano banana generate images, low as $0.04 per image
by•
Nano Banana (or Gemini-2.5-Flash-image) is a strongly consistent image generation and editing model that outperforms Flux Kontext. It is Free, Stable, and Fast.
Replies
Best
Maker
📌
Just launched Nano Bananas 🎉 — an AI image tool that’s super affordable (as low as $0.04 per image!) but still delivers top-tier quality.
What really impressed me:
• Consistent characters — no more random faces changing between edits.
• No prompts needed — you can instantly turn a photo into a figure/figurine-style render with one click.
• 16 emoji pack in seconds — generate a full set of expressions from just one photo.
• Seamless fusion — up to 12 images blended into one, with zero awkwardness.
For me, the mix of low cost + high quality makes it a real game-changer compared to other AI editors. Curious to hear what kind of creative uses people will come up with here! 🚀
Report
Maker
2025-10-09
Added password login feature - users can now set a password and login using email + password
Premium users now support dual image generation, advanced users can generate up to 32 images simultaneously
2025-10-08
Sora2 model now supports video duration selection: 10 seconds (default) and 15 seconds
Optimized video generation interface layout, adjusted prompt area height to accommodate duration selector
Dynamic credit calculation based on video duration and resolution (10s 720p=200 credits, 15s 720p=1000 credits; 10s 1080p=1000 credits, 15s 1080p=1200 credits)
Removed moment.js dependency, using lightweight custom date utilities, reducing bundle size by ~288KB
Optimized code splitting and build configuration for faster page loading
2025-10-03
Launched new Sora2 video generation model with improved quality
Reduced Veo3 video generation pricing from 800 to 150 credits, significantly lowering the barrier to entry
Optimized mobile (WAP) device page adaptation for better display
Enhanced stability of video generation models
Report
Maker
Veo3.1 Online
My feelings about Veo 3.1 are very mixed. It didn't bring the expected "disruptive" revolution, but in some ways it's incredibly powerful. I have two conclusions: (1) It's weak in narrative and creative transitions. (2) The more your prompt resembles a movie script, the more amazing the effect.
Honestly, the Veo 3.1's video capabilities gave me a very strange experience. It's a bit biased.
The Veo 3.1's physics simulation is incredibly powerful, a true step up from the Veo 3. It simulates light and shadow, gravity, and even the movement of fabric, and it does so with remarkable realism. The lip-syncing capabilities are also significantly improved, and the background music sounds are incredibly well-matched!
Replies
2025-10-09
Added password login feature - users can now set a password and login using email + password
Premium users now support dual image generation, advanced users can generate up to 32 images simultaneously
2025-10-08
Sora2 model now supports video duration selection: 10 seconds (default) and 15 seconds
Optimized video generation interface layout, adjusted prompt area height to accommodate duration selector
Dynamic credit calculation based on video duration and resolution (10s 720p=200 credits, 15s 720p=1000 credits; 10s 1080p=1000 credits, 15s 1080p=1200 credits)
Removed moment.js dependency, using lightweight custom date utilities, reducing bundle size by ~288KB
Optimized code splitting and build configuration for faster page loading
2025-10-03
Launched new Sora2 video generation model with improved quality
Reduced Veo3 video generation pricing from 800 to 150 credits, significantly lowering the barrier to entry
Optimized mobile (WAP) device page adaptation for better display
Enhanced stability of video generation models
Veo3.1 Online
My feelings about Veo 3.1 are very mixed. It didn't bring the expected "disruptive" revolution, but in some ways it's incredibly powerful. I have two conclusions: (1) It's weak in narrative and creative transitions. (2) The more your prompt resembles a movie script, the more amazing the effect.
Honestly, the Veo 3.1's video capabilities gave me a very strange experience. It's a bit biased.
The Veo 3.1's physics simulation is incredibly powerful, a true step up from the Veo 3. It simulates light and shadow, gravity, and even the movement of fabric, and it does so with remarkable realism. The lip-syncing capabilities are also significantly improved, and the background music sounds are incredibly well-matched!