Gemini AI Navratri Photo Trend

Lisa Ernst · 09.11.2025 · Technology · 6 min

The creation of Garba- and Dandiya- looks from selfies using artificial intelligence, especially with Google Gemini, connects traditional festival culture with digital creativity. This trend, which developed around the Hindu festival Navratri, enables users to present themselves in elaborate outfits and festive scenes. Precise prompts play a key role, while questions about data privacy, cultural sensitivity, and the limits of technology arise.

Source: Own portrayal

Introduction & Context

Navratri is a nine-day Hindu festival in honor of the goddess Durga, whose name is derived from the Sanskrit words 'nava' (nine) and 'ratri' (night). It is celebrated in many regions of India with dancing, prayer, and festivities, especially in the western part of the country. Garba, a circle dance from Gujarat, and Dandiya Raas, , a dance with decorative wooden sticks, are central components of these celebrations and have also spread in the global diaspora. Typical clothing for women includes multi-piece outfits such as ghagra or chaniya (long skirt), choli (short top) and dupatta (shoulder scarf), often richly embroidered with mirror work, Bandhani tie-dye and plenty of jewelry such as bangles, necklaces, and nose rings.

Google Gemini is an AI system that can edit text and images. The image model 'Gemini 2.5 Flash Image', known as 'Nano Banana', enables users to transform selfies with text instructions into new scenes and looks. A Navratri- or Garba look with Gemini means editing a real selfie so that the person appears in a Garba scene, in a Chaniya-Choli outfit or in a Dandiya night, often in 4K quality with cinematic lighting and film grain.

Current Trend & Use

The autumn Navratri 2025 fell during the period from September 22 to October 2, during which many of these AI photos appeared on social media. Media in India described how Gemini, with the Nano-Banana image model, transforms selfies into festive Garba and Dandiya scenes, including detailed outfits and lighting moods. Beispielhafte Prompts which include scenes of people in embroidered Ghagra-Choli combinations that swirl in dance, holding Dandiya sticks, with warm evening light and retro film grain. Other suggestions dress selfies in 'royal' Rajasthan looks with palace backdrops or black, mirror-embellished lehengas with silver jewelry.

The Nutzung ist einfach: Download the Gemini app, sign in with your Google account, upload a selfie, enter a prompt, and generate the finished portrait. It is recommended to explicitly mention lighting, background, and 'film grain' in the prompt to achieve realistic results. In parallel, there are other photo formats such as Retro Saree portraits and 3D figures. Gemini gained over 23 million new users in a few weeks with the Nano Banana feature and enabled more than 500 million image edits, with India being an important market.

Source: oneindia.com

The 'Gemini AI Navratri Photo Trend' enables the creation of impressive images like this that capture the festive mood and traditional clothing.

Motives & Interests

Users want to present themselves during the festival in the way their ideal Garba or Dandiya fantasies dictate, without access to expensive outfits or professional photography. For people who live outside India or cannot attend large Garba evenings, AI becomes a digital stage to connect with tradition. For Google this is strategically interesting, as image editing with Gemini increases time spent in the app and makes the product emotionally tangible for a mass market. Viral photo formats act as free advertising, as users share their pictures and others download the app.

Tutorial videos and guides reinforce the trend by offering step-by-step instructions for creating Garba night scenes. Many guides recommend similarly structured prompts: upload a clear selfie, precisely describe the Ghagra- or Chaniya-Choli outfit, add jewelry, define the background, specify lighting mood and desired visual appearance. At the same time, power and data issues come into play: those who upload selfies provide biometric information and personal details that can be stored and analyzed for training or security purposes. Experts warn that image generators can create convincingly realistic deepfakes from only a few photos, and unclear deletion policies as well as third-party apps add additional risks.

Risks & Concerns

It is established that Google Gemini with 'Nano Banana' transforms selfies into new image styles and allows up to 100 free edits per day. The increased search activity for 'Google Gemini Photo' and related terms, especially in Indian states like Gujarat, is also documented. However, it remains unclear how stable and fair the image generation is for all skin tones and face shapes. Reports suggest that AI models may tend to brighten skin or reinforce certain beauty ideals, while scientific studies show general biases toward lighter skin tones. Systematic studies on this Navratri-specific application of Gemini do not yet exist.

The assumption that the AI Navratri images are an official Google campaign intended to replace traditional rituals is false; the initiative mostly comes from users, media, and influencers. Also misleading is the idea that AI images are harmless. Privacy and legal experts point to real risks such as data leakage, deepfakes, and fraud attempts, especially when using unofficial websites. An Indian police officer publicly warned about fraudulent websites that pose as free AI tools to collect personal data and photos. He recommends using only the official Gemini app.

Source: freejobalert.com

Not only women, but also men use AI tools to present themselves in festive Navratri clothing against traditional backdrops, such as here in front of Durga Puja decorations.

Practical Tips

Whoever wants to create their own Navratri or Garba looks with Gemini should use the legitimate Gemini app via Google Play or the Apple App Store and sign in with their Google account. It is recommended not to upload particularly intimate or sensitive photos and to regularly review privacy options. For good results, a precise prompt is important. Terms such as Ghagra, Choli or Chaniya-Choli, fabrics (e.g., Bandhani, mirror work), jewelry (oxidized bangles, nose ring, necklaces), lighting ("golden hour", "soft studio lighting") and background ("Garba ground with dancers in a circle", "Dandiya Night with many lights") should be described in detail. Also the desired camera angle (e.g., "Full-body portrait", "Mid-Twirl"), the aspect ratio (e.g., "4:5 vertical"), and a light cinematic grain contribute to a cohesive look.

An example prompt could look like:\n"Full-body portrait of a woman in a vibrant red Ghagra-Choli with intricate mirror work, dancing gracefully at a Navratri Garba night. She holds Dandiya sticks, adorned with oxidized silver jewelry, including bangles and a nose ring. Warm golden hour lighting, cinematic film grain, blurred background with other dancers in a circle. Vertical 4:5 aspect ratio, maintaining original skin tone."

Particularly with skin tones and facial features it makes sense to explicitly state in the prompt that one's own skin tone should be preserved and the face should remain recognizable, as many AI systems tend to brighten or smooth skin. Culturally respectful use is encouraged if you also inform yourself about the background of Navratri, Garba, and Dandiya, as these dances are deeply rooted in religious and communal practices.

Source: user-added

Young men use the Gemini AI Navratri photo trend to stage themselves in front of representations of the Goddess Durga and express their devotion.

Outlook & Open Questions

It remains open how strongly AI Navratri images will change the experience of the festival in the long term. To date, there are mainly journalistic observations and data points on search queries, but no comprehensive social science studies on whether people attend Garba evenings physically less because of AI photos. It also remains unclear how strongly systemic biases in training data—such as regarding skin tones or body shapes—shape the generated Navratri images. Research on AI image systems shows that darker skin tones are underrepresented in some datasets and certain beauty standards can be reinforced.

Finally, the legal aspect remains in motion. Articles from tech and legal contexts discuss privacy, potential liability for deepfakes, and the question of how long corporations like Google store images and what they may be used for. National and international regulations are still being developed, and it is not yet clear how strict these rules will be in the future and how well they will protect festival and portrait images in particular.