Google Unveils Nano Banana 2.1: Superior Image Generation at a Fraction of the Cost

In brief

• Google released Nano Banana 2.1 on October 6, rolling it out across the Gemini app, Google Search’s AI Mode, Google Ads, and its developer tools.
• Internal evaluations indicate the new model outperforms both Nano Banana 2 and Nano Banana Pro, achieving an infographic factuality score of 0.521 compared to 0.179 for its predecessor, though independent benchmarks are pending.
• Developer pricing starts at $0.0336 for a standard 1K image, roughly half the cost of the previous version.

Google launched Nano Banana 2.1 on Tuesday, introducing its latest AI model designed to transform text prompts into images and perform advanced photo editing.

The update is being deployed across key platforms, including the Gemini app, Google Search’s AI Mode, Google Ads, and developer tools such as Google AI Studio, Flow, and Stitch.

The original Nano Banana model was a major driver for Gemini’s popularity, famously leveraging selfie-to-figurine generation to top both major app stores in September 2025, dethroning ChatGPT after nearly three years.

During that same period, Alphabet’s market valuation surpassed the $3 trillion milestone.

Version 2 launched in February, built on the Gemini 3.1 Flash Image model. Its key innovation was integrating Google Search to verify real-world facts, resulting in more accurate depictions of actual events and individuals.

Meet Nano Banana 2.1, our latest image generation and editing model

This upgraded version outperforms our previous models across the board, with notable leaps in visual design, mask-based editing, and subject consistency to help you create more natural-looking images. pic.twitter.com/d858DrzrF3

— Google (@Google) October 6, 2026

Google highlights three core upgrades in this release: enhanced visual design, more precise mask-based editing (allowing localized changes to specific regions), and improved subject consistency to ensure individuals and objects remain recognizable across multiple modifications.

In overall text-to-image preference testing, Nano Banana 2.1 achieved 1,050 ELO points, outperforming Nano Banana 2 (990 points) and Nano Banana Pro (935 points). In this qualitative, non-capped ELO rating system, higher scores represent superior user preference.

Technically, the model supports up to 14 reference images simultaneously, tracking up to four characters and ten distinct objects. It generates outputs up to 4K resolution with aspect ratios as wide as 8:1.

Developers can adjust the model’s reasoning time from minimal to high, and enable grounding via Google Search and Google Image Search, allowing the AI to retrieve and integrate real-world information before generating images.

Via Google’s developer API, which enables third-party applications to access the model, a standard 1K image is priced at $0.0336—approximately half the $0.067 cost of Nano Banana 2.

4K image generation is priced at $0.0756 compared to $0.151, with an additional 50% discount available for batch processing jobs.

At the standard 1K rate, producing one thousand images costs approximately $33.60, down from $67 with the previous model, making it an exceptionally efficient and cost-effective option.

Source link