Introduction
An Indian Girl Era GRWM workflow is an AI pipeline that produces a vertical get ready with me video in the warm desi editorial aesthetic, complete with Hinglish voiceover and ambient Indian instrumentation. The workflow uses a Devanagari mixed script method to fix Hindi pronunciation issues in ElevenLabs. This guide shows the full eight node Vertex build, prompts, and a hero image prompt you can run standalone.
What Is the Indian Girl Era Aesthetic
The Indian Girl Era aesthetic blends modern editorial framing with traditional desi visual vocabulary. Think warm saffron and rose tones, filigree jali light patterns, marigolds, brass diyas, fine mirror work on cotton kurtas, stacked gold bangles, and a soft makeup look with kohl and berry lips. Reference points include Sabyasachi campaign editorials and contemporary Indian fashion photography. Mood is warm, grounded, confident.
What the Workflow Produces
A 15 to 20 second vertical 9:16 video with a warm cinematic still of the protagonist, a one shot gentle camera push or drift, a Hinglish voiceover with proper Hindi pronunciation using the Devanagari mixed script method, and a light Indian instrumental bed. Output is Instagram Reels and Takatak ready.
Hero Image Prompt You Can Run Today
Run this prompt directly in Flux Kontext, Midjourney v7, or Nano Banana to generate the hero image that accompanies this blog post. The prompt is tuned for a 9:16 vertical composition and can be adapted to 1:1 for feed posts.
A young Indian woman in her mid twenties, warm brown skin with a natural glow, long wavy dark hair with a side part, glossy kohl lined eyes, a soft berry lip, a tiny diamond nose pin. She wears a soft peach cotton kurta with fine mirror work, stacks of thin gold bangles on her wrist, small gold jhumkas. She sits at a low wooden carved dressing table with small brass diyas, a silver kajal pot, fresh marigold petals scattered on a muslin cloth, and a small mirror catching soft morning light. The setting is a warm Mumbai apartment with filigree jali windows casting soft patterned light across her cheek. Natural golden hour light pours from camera right. Shot on 35mm film, Kodak Portra 800 grain, shallow depth of field at f 1.8, warm saffron and rose color grade, editorial mood similar to Tarun Khiwal photography. Vertical 9:16 composition.
The Devanagari Mixed Script Method
ElevenLabs multilingual v2 mispronounces Hindi words when written in Latin script. The fix is to write Hindi words in Devanagari script while keeping English surrounding text in Latin. Example script: Hi guys, aaj main share kar rahi hoon my morning रूटीन. Starting with थोड़ा सा warm water और फिर skincare. This method is now the MinionArts standard for any Hinglish voiceover work.
Step by Step Vertex Build
The workflow uses eight nodes: an interface form, a CHAT node for the visual prompt, a Flux Kontext keyframe, a Veo animation, a TEXT_TO_SPEECH node with the Devanagari prepared script, a music generator, a TEXT_JOINER to prepare the audio merge, and a final VIDEO_AUDIO_MERGE.
Node Build Table
Build these nodes in order on the Vertex canvas. Each node below lists its execution mode and key fields with wiring details.
Step 1: node_interface_form
Execution Mode: INTERFACE_FORM
Fields and Wiring: node_type INTERFACE. Outputs: creator_description (TEXT), setting_detail (TEXT), hinglish_script_devanagari (TEXTAREA), aspect_ratio (ASPECT_RATIO default NINE_SIXTEEN)
Step 2: node_chat_visual_prompt
Execution Mode: CHAT
Fields and Wiring: System instruction generates a Flux Kontext image prompt from creator_description and setting_detail. Output: generated_text
Step 3: node_flux_desi_keyframe
Execution Mode: CREATE
Fields and Wiring: Inputs: prompt from chat node via TEXT_JOINER with anchor style block, aspect_ratio from interface. Output: generated_asset IMAGE
Step 4: node_veo_animate
Execution Mode: VIDEO
Fields and Wiring: Inputs: reference_image_1 from flux output, prompt static hold with subtle breeze on dupatta 8 seconds, aspect_ratio NINE_SIXTEEN. Output: generated_asset VIDEO
Step 5: node_tts_hinglish
Execution Mode: TEXT_TO_SPEECH
Fields and Wiring: Inputs: text_input from interface_form.hinglish_script_devanagari. voice_id set to warm Indian female voice. Output: generated_audio SPEECH
Step 6: node_music_indian
Execution Mode: MUSIC_GENERATOR
Fields and Wiring: Prompt: soft sitar and tanpura with light tabla 85 BPM warm contemplative morning Indian instrumental 15 seconds. Output: generated_audio AUDIO
Step 7: node_compose_audio
Execution Mode: COMPOSE
Fields and Wiring: Combines TTS speech and music bed into one track. Output: composed_video (used for audio track prep)
Step 8: node_final_merge
Execution Mode: VIDEO_AUDIO_MERGE
Fields and Wiring: Inputs: video_reference from veo, audio_reference from compose. Output: merged_video
Flux Kontext Keyframe Prompt
Anchor prompt for node_flux_desi_keyframe: young Indian woman mid twenties, warm brown skin natural glow, long wavy dark hair, kohl lined eyes berry lip small diamond nose pin, soft peach cotton kurta with mirror work, stacked thin gold bangles small jhumkas, low carved wooden dressing table, brass diyas silver kajal pot marigold petals on muslin, warm Mumbai apartment with filigree jali window casting patterned light, 35mm film Kodak Portra 800, shallow depth of field f 1.8, saffron and rose color grade, editorial mood, vertical 9:16.
Veo Motion Prompt
Motion prompt for node_veo_animate: subject holds still gaze direction constant, subtle breeze moves dupatta edge and hair strands, gentle camera drift from right to left at half speed, jali light pattern slowly shifts across cheek, single unbroken take eight seconds, warm film grain consistent throughout.
FAQ
Why Devanagari mixed script instead of full Latin script?
ElevenLabs multilingual v2 reads Hindi words written in Latin as English phonemes which produces incorrect pronunciation. Writing the Hindi words in Devanagari script triggers the correct Hindi pronunciation pathway. Surrounding English stays in Latin so accent transitions feel natural.
Can I use this for non Hinglish voiceover?
Yes. Replace the TTS script with pure English or pure Hindi depending on the target audience. The visual aesthetic holds independent of language choice.
What aspect ratio options are supported?
The interface form defaults to NINE_SIXTEEN for Reels. You can also set SIXTEEN_NINE for YouTube shorts horizontal or ONE_ONE for feed posts
