← all prompts

Turn your photo into a Radha–Janmashtami portrait

portrait ChatGPT (GPT-4o image) Updated 4 Sept 2026

Your face stays exactly as your photo. Everything else becomes a Radha-inspired Janmashtami portrait — pink and silver, a jasmine gajra, and Krishna's blue hand resting at your chin.

Two runs, one prompt

Radha–Janmashtami portrait with Krishna's blue hand at the chin
The signature frame — blue hand at the chin, eyes closed.
Radha–Janmashtami portrait in warm candlelight
Same prompt, candlelit and hands folded — ask for the variation you want.

Both came from the one prompt below. Ask ChatGPT for other poses and angles, then stitch them into a reel.

See it in motion

Portraits and reel by @sujata__behera.

The prompt

Paste into ChatGPT with your photo

Create an ultra-realistic cinematic vertical 9:16 portrait of the same woman from the uploaded reference image. Use the uploaded photo as the ONLY facial identity reference. Transform her into an enchanting Radha-inspired Krishna Janmashtami look. She is wearing an elegant pink traditional outfit richly decorated with delicate silver and mirror-work embroidery, layered pearl necklaces, ornate floral earrings, and a large traditional kundan maang tikka. Her black hair is styled in a graceful half-up traditional hairstyle, beautifully surrounded by a voluminous gajra of tiny white jasmine flowers, with a few soft strands framing her face. Her makeup is soft and divine: shimmery rose-pink eyeshadow, long defined eyelashes, subtle eyeliner, natural glowing skin, rosy cheeks, and rich pink lips. Add intricate white Radha-style facial adornments around the eyebrows and forehead, with a beautiful white U-shaped Vaishnav tilak and a tiny red decorative dot at the center. She has her eyes gently closed with a soft, innocent smile, creating a peaceful devotional expression. A graceful blue Krishna-inspired hand/glove gently cups her chin, while her own decorated hands rest elegantly near her face, adorned with delicate rings, floral jewelry, and pearl strands. Use dreamy studio lighting, soft diffused highlights, subtle blue and warm golden rim light, creamy bokeh background, shallow depth of field, premium DSLR photography, 85mm portrait lens, f/1.8, realistic skin texture, intricate jewelry details, cinematic color grading, luxurious devotional atmosphere, photorealistic, ultra-detailed, 8K quality. No artificial/AI-looking face, no facial distortion, no extra fingers, no duplicate jewelry, no text, no watermark. Match the face features to the reference photo.

Written for one person and a 9:16 crop — add “square” or “4:5 crop” if you want a feed post instead.

How to run it

  1. Pick the photo. A clear, front-facing photo of just you, evenly lit — this is the only face reference the model gets.

  2. Upload and paste. Attach the photo in ChatGPT and paste the prompt.

  3. Ask for more frames. Once you like the look, ask for the same woman, same styling in another pose or angle — hands in namaste, looking up, a soft profile — one at a time.

For a better result

On likeness: AI keeps faces close, not perfect. Treat the result as a stylised portrait, not a photograph.