Google's Gemini Nano Banana 2.1 Enters Arena Image Leaderboards — Thinking-Capable with C2PA Credentials
Google added Gemini Nano Banana 2.1 to Arena’s Text-to-Image and Image Edit leaderboards on October 6, opening the model to head-to-head evaluation against competitors including grok-imagine-image-2.0 (canvas), which joined those same leaderboards four days earlier.
Nano Banana 2.1 is part of Google’s Gemini Enterprise Agent Platform and is optimized specifically for image generation and editing, not general-purpose text tasks. The model ID is gemini-nano-banana-2.1; it is accessible via Agent Studio and the Model Garden on Google Cloud.
Capabilities
| Feature | Status |
|---|---|
| Image generation | Supported |
| Image generation from video input | Supported |
| Multi-turn image editing | Supported |
| Interleaved images and text | Supported |
| Virtual try-on | Supported |
| Thinking | Supported |
| Content Credentials (C2PA) | Supported |
| Person generation | Supported |
| Google Search grounding | Supported |
| Image search grounding | Supported |
Thinking in an image model is the headline capability. Most text-to-image systems do not expose a reasoning layer; Nano Banana 2.1 includes it, positioning the model for complex multi-step generation requests that benefit from planning before pixel output.
C2PA content credentials embed cryptographic provenance metadata into generated images, allowing downstream tools to verify that an image was AI-generated and identify the originating model. With AI-generated images increasingly appearing in publishing and social media, built-in attribution is a differentiator.
Context Window
- Context: 131,072 tokens
- Max output tokens: 32,768
- Implicit context caching: supported
What It Does Not Do
The model does not support function calling, code execution, the Interactions API, Gemini Live API, structured output, RAG Engine, or fine-tuning. It is a specialist model, not a general-purpose agent.
Arena Position
The leaderboard addition opens direct comparative evaluation against other frontier image generators. grok-imagine-image-2.0 (canvas) and Claude Sonnet 5.5 (added to Text and Vision Arena on October 2) were the other recent October additions before Nano Banana 2.1 arrived. Head-to-head ELO scores will accrue as Arena users evaluate the model in Text-to-Image and Image Edit battles.