Google Enhances Gemini AI Image Capabilities to Challenge ChatGPT

Google has enhanced the image generation capabilities of its Gemini AI model, introducing a new version known as Gemini 2.5 Flash Image. The update is seen as a direct competitive response to OpenAI’s ChatGPT, aiming to bridge the gap between the two platforms by offering more precise and consistent image generation and editing features. The model can now generate high-quality images with greater accuracy, maintain character consistency across scenes, and allow users to manipulate visuals using natural language commands—such as adjusting poses, merging images, and changing lighting—all while preserving the integrity of faces and environments [1].

Developers can now access the updated tool through Google’s AI Studio, as well as via third-party platforms like OpenRouter and fal.ai, which are expanding the model’s availability to a wider range of coders and AI enthusiasts. The model is available in preview mode and is being marketed under the informal name “nano-banana” on LMArena, a crowdsourced AI testing site. The tool is particularly praised for its seamless editing capabilities and ability to interpret complex prompts that combine text and visual references [1].

The model carries a cost of $30 per million output tokens, or roughly four cents per image, through Google Cloud. To address concerns around AI-generated content misuse, all outputs will be marked with an invisible SynthID watermark and metadata tag to indicate their artificial origin [1].

OpenAI remains a formidable competitor in this space, having introduced image generation capabilities in March 2025 with its GPT-4o model, which contributed to ChatGPT surpassing 700 million weekly active users. In comparison, Google reported 400 million monthly active users for Gemini in August 2025, suggesting a significant but not insurmountable gap. The new image capabilities aim to not only attract users with enhanced functionality but also to reinforce Google’s position in the broader AI arms race [1].

Analysts suggest that the timing of the Gemini 2.5 Flash Image release is strategic, following a period of intense competition and regulatory scrutiny in the AI space. The broader AI ecosystem is shifting toward multimodal capabilities, with companies like Perplexity and OpenAI introducing tools that combine image generation with real-time research and interactive web experiences. By improving Gemini’s image capabilities, Google is aligning itself with the evolving expectations of users who demand more integrated and versatile AI tools [1].

The continued development of AI image generation also reflects an industry-wide push toward making AI more accessible and practical in creative and professional workflows. However, it also raises ongoing concerns about transparency, authenticity, and the potential for misuse—issues that Google and other major players are actively addressing through watermarking and metadata tagging [1].

Source: [1] https://decrypt.co/336878/google-boosts-gemini-ai-capabilities-latest-salvo-against-chatgpt