Nano Banana 2 : la méthode simple pour créer des images au TOP

Nano Banana 2 : la méthode simple pour créer des images au TOP

by Ludo Salenne

Applied Sciences & Engineering AI — Models & Research UYQArtificial intelligence (AI)UYQFGenerative artificial intelligence
🎙 Ludo Salenne 👥 266K 📅 November 20, 2025 ⏱ 23 min 👁 36K 📄 tutorial 🧭 2026-08-02
Available in: English (current) Français

Keywords

Nano Banana ProGoogle Geminiimage generationAI tutorialprompt methodology

Summary

In this video, Ludo Salenne provides a comprehensive tutorial on Google’s new AI image generation model, Nano Banana Pro (also known as Nano Banana 2). He demonstrates its capabilities through various tests, including generating photorealistic images, editing photos with simple drawings, transforming sketches into product renders, and creating promotional posters. He highlights the model’s ability to handle text rendering and its integration with Google Gemini 3 Pro. Salenne also introduces his ‘CRYSTAL’ prompt methodology, which he claims to have developed based on training from Google, and shows how to use a custom Gem to generate optimal prompts. The video covers advanced features such as combining multiple images, converting videos into images, and using Canvas for iterative editing. Throughout, he emphasizes the speed and quality of Nano Banana Pro, positioning it as a leading tool in AI image generation. The tutorial is practical and aimed at creators looking to leverage AI for visual content.

153 words

Critical Evaluation

The video offers a practical, hands-on demonstration of Nano Banana Pro’s capabilities, which is valuable for users interested in AI image generation. The creator’s early access and direct testing provide authentic insights into the model’s performance, including its strengths in text rendering and photorealistic output. The introduction of the CRYSTAL prompt methodology, while not deeply explained, suggests a structured approach to prompt engineering that could be beneficial. However, the video lacks critical analysis of limitations, such as potential biases or ethical concerns, and the evaluation is largely subjective, based on personal preference. The sources cited are primarily the creator’s own resources and tutorials, which may not provide independent verification. The ad integration is present but not disruptive. Overall, the content is informative and well-presented, but it would benefit from more balanced scrutiny and references to external research or benchmarks.

139 words

Title / Content Match

The title accurately reflects the content, which focuses on demonstrating the capabilities and usage of Nano Banana 2 (Pro) for image generation.

Quality & Reliability

7/10

The video is a hands-on tutorial by a content creator with early access to Nano Banana Pro. It demonstrates practical usage and provides a prompt methodology (CRYSTAL) based on training from Google. However, it lacks in-depth technical analysis or independent verification, and relies on personal experience and promotional content.

Chapters

Cited Sources

Concurring Sources

  • Google Gemini — Official product page for Google's AI model, which includes Nano Banana Pro.

External References

Contribution & Novelties

The video provides a practical overview of Nano Banana Pro’s features, including the ability to edit images with drawings, generate consistent characters, and combine multiple images. It introduces the CRYSTAL prompt methodology, which is a structured approach to crafting prompts for optimal results. The demonstration of using a custom Gem to automate prompt generation is innovative for streamlining workflows.

Pour aller plus loin :

  • Google Gemini — Official page for Google’s AI model, relevant for understanding the platform.
  • Prompt Engineering Guide — Comprehensive guide on prompt engineering techniques, relevant to the CRYSTAL methodology.
  • Diffusion Models — Background on the underlying technology of AI image generation.

105 words

Radar Profile

The radar profile shows high scores in quantity of information and reliability, reflecting the video's comprehensive coverage and the creator's early access. The technical level is moderate, indicating accessibility for a broad audience. The overall balance suggests a well-rounded tutorial with practical insights.

Reliability 7/10

💬 Très positif : Sur les 30 commentaires analysés, les spectateurs expriment une forte appréciation, saluant la pédagogie et la réactivité du créateur, avec des encouragements et des questions techniques.