A navratri ai photo editing prompt is a long text prompt pasted into Gemini or ChatGPT with a selfie to get a garba or dandiya night portrait in festive dress. On videosi you skip the prompt: upload one photo and get one vertical 2:3 image.
This guide goes past the basics on the trend page. It covers boy and girl outfits, why copied prompts change your face, how to shoot the source photo, couples, shrines, sizes and next year's dates.
Key takeaways
- Women get a chaniya choli and men get a kediyu, and the outfit has to be named so the model does not drift to a generic lehenga.
- Copied prompts change your face when they praise a look instead of locking the identity to your photo.
- The couple version keeps two adults side by side, a step apart, with no hugging or kissing.
- The result is a silent 2:3 still image, not a video, and it costs 1 credit.
- Navratri 2026 ran from 11 to 19 October, so this page is the reference for next year.
In this guide
- How is a Navratri AI photo editing prompt for a boy different from one for a girl?
- What is the difference between a chaniya choli and a kediyu?
- Why does a copied Navratri prompt give me a face that is not mine?
- How should I shoot the original selfie?
- Why does the Navratri couple photo show two people side by side and not close together?
- Why does the photo not include a goddess, an idol or a shrine?
- Which scene and pose should I pick?
- What size is the image, and is it good for a WhatsApp status?
- When is Navratri next year, and when should I make the photo?
- What is garba, and how is it different from dandiya raas?
- What does a Navratri AI photo cost, and what do I get?
How is a Navratri AI photo editing prompt for a boy different from one for a girl?
The outfit is the main difference, and it changes more than you would expect. For a woman the prompt asks for a flared chaniya choli with a ghagra skirt, a short blouse, a sheer dupatta, jhumka earrings and bangles. For a man it asks for a short flared kediyu jacket with mirror-work, a matching kurta, dhoti pants and a bandhani stole, and it should add no jewellery beyond what the selfie already shows. A prompt written only for a girl and pasted over a man's selfie usually gives a skirt, a long hair line or a softened jaw, which is the face drifting toward the outfit.
The videosi page reads the outfit from the person in the photo, so a man gets the kediyu and a woman gets the chaniya choli without you editing any wording. Posts in this trend mostly show girls, so there are fewer boy examples to compare against, and it is worth trying both poses before you decide.
What is the difference between a chaniya choli and a kediyu?
A chaniya choli is the women's garba outfit: a wide, flared skirt (the chaniya or ghagra), a fitted short blouse (the choli) and a dupatta. Bright colours, round mirror-work and thread embroidery are the signature. A kediyu is the men's counterpart: a short jacket that flares at the hem, worn over a kurta and dhoti-style pants, often with a bandhani stole.
This matters for AI because the two outfits move differently. The skirt of a chaniya choli makes a big circle in a twirl, while a kediyu flares in a smaller, sharper way. If your prompt says only "traditional dress", the model tends to fall back on a generic lehenga or saree. Naming the garment, the mirror-work and the colour is what keeps the result looking like a garba night and not a wedding.
Why does a copied Navratri prompt give me a face that is not mine?
Most copied prompts describe a look, not a person. They say things like "beautiful face, glowing skin, perfect smile", and the model obeys by producing a more glamorous stranger. Others never say which part of the picture must stay fixed, so the model may rebuild the face along with the outfit.
The fix is a rule written first: the face, age, skin tone, build and hair come from the uploaded photo, and only the clothes, the pose and the setting change. Beauty words and slimming words should be left out completely. This is the identity lock that the videosi Navratri page is built around. Even with it, likeness is not guaranteed, because new lighting and a new outfit can shift small details, so judge the result on your first try and adjust the photo before you change the wording.
How should I shoot the original selfie?
Use one person, facing the camera, in even soft light, with the face and shoulders in frame. A window or open shade works better than a flash. Take off sunglasses, and skip beauty filters that smooth the skin, because the model copies whatever it sees, including the smoothing.
You do not need a full-length photo, since the outfit is replaced. A sharp face matters more than what you are wearing. Loose hair suits the garba twirl, while any hairstyle works for the dandiya pose. If your only good photo is a group shot, crop it down to one face before you upload it. A small face in a crowd is one of the most common reasons the result does not look like the person.
Why does the Navratri couple photo show two people side by side and not close together?
Because prompts that ask for closeness tend to merge the two faces. When a single prompt says "couple" or adds a romantic mood, the model pulls the heads together and blends features, so each person ends up looking like a mix of both. Keeping the two dancers a small step apart, each facing the camera with their own pair of dandiya sticks, avoids that.
It also matches the dance. Dandiya raas is danced in pairs, and according to the Wikipedia article on Dandiya Raas, dancers typically stand in two facing lines and strike sticks with a partner before moving on to the next one. Two people dancing side by side is closer to the real thing than an embrace. On videosi the couple version takes one photo of each person, and each face comes only from its own photo. Both are adults, and there is no hugging or kissing.
Why does the photo not include a goddess, an idol or a shrine?
Navratri is a religious festival, and a prompt that asks for the goddess or a puja backdrop has two problems. Image tools often block or distort religious figures, and when they do render a shrine it is large enough to push you out of the frame. The result stops being your portrait.
So the garba night photo keeps the festive feeling through light and colour instead: string bulbs, paper lanterns, rows of clay diyas and marigold garlands, with blurred dancers behind you. The trend page states the rule plainly: adults only, and no deities, idols or temple shrines anywhere in the frame.
Which scene and pose should I pick?
There are three scenes and two poses, so six combinations. The garba ground under round string bulbs is the classic and the safest. The moonlit courtyard with diyas on the floor looks the most cinematic. The street garba under festoon lights and marigold garlands has the most crowd energy.
For poses, the dandiya pose (one painted stick raised high, one at chest height, a big smile) shows your face and outfit clearly and is the better choice when likeness matters most. The garba twirl is a full-body spin with the skirt or kediyu flaring and the dupatta or stole flying, which is more dramatic but gives the face less room. If you are unsure, start with the dandiya pose on the garba ground and change one thing at a time.
What size is the image, and is it good for a WhatsApp status?
The result is one still image in vertical 2:3. That ratio is tall enough to fill most of a phone screen, so it suits a WhatsApp status or an Instagram story, and it also works as an ordinary feed post.
A 2:3 picture is a little wider than the 9:16 of a full-screen story, so some apps will show thin bars at the top and bottom or crop the edges slightly. Keep that in mind and you will not be surprised. The image has no sound, so if you post it as a Reel, add your own garba song in the app. Any other music is the creator's choice, since there is no single sound for this trend.
When is Navratri next year, and when should I make the photo?
Navratri moves with the Hindu lunisolar calendar, so the Gregorian dates change every year. The Wikipedia article on Navaratri lists Sharada Navaratri in 2026 as 11 October (Sunday) to 19 October (Monday), nine nights in the autumn. That festival has now passed.
The 2027 dates are not confirmed here, so check a calendar when the time comes. In practice the posts appear a few weeks before the first night, and the page stays up as the reference for the next season. If you want the picture for a profile photo or a greeting, make it a week or two early so you can try more than one scene.
What is garba, and how is it different from dandiya raas?
The Wikipedia disambiguation page on Garba describes it as a form of dance originating from Gujarat, India, and the Navaratri article says Gujarati celebrations consist of arti followed by garba. Garba is danced in circles, mostly with hands and footwork. Dandiya raas, the same festival's stick dance, is performed with a pair of wooden sticks, and the Dandiya Raas article says it is danced during Sharad Navratri as well as at Janmashtami fairs and weddings.
That is why the photo offers both. The dandiya pose holds the sticks up, and the twirl is closer to a garba spin. If you are searching for ideas, both words lead to the same kind of picture, so use whichever pose reads better in your photo.
What does a Navratri AI photo cost, and what do I get?
Each image costs 1 credit. Credits are sold in packs, and the Single pack is $6.99 for 10 credits, which is 10 pictures. In India the Single pack is $3.99. Credits do not expire, and a failed or blocked generation returns its credit automatically. It is a paid tool, not a giveaway.
You get one vertical 2:3 still image, usually in about a minute, from one or two photos. It is made by OpenAI's GPT Image 2.5 inside the Navratri page, with no prompt to write or paste. videosi is independent and not affiliated with Google, OpenAI or any social app.
