Facetra is an AI-powered jewellery design workflow platform for Indian ateliers. It converts hand sketches into photorealistic renders, multi-view technical sheets, live-gold-rate Bill of Materials, and downloadable 3D meshes (GLB/STL/OBJ/FBX). Built for Jadau, Polki, Kundan, Meenakari, Temple and Rajwadi styles.

Does Facetra replace CAD software like Rhino or MatrixGold?

No. Facetra is a workflow layer that produces the reference material (renders, multi-view sheets, BOM, 3D mesh) a CAD operator uses. Your CAD bench continues to use Rhino, MatrixGold or JewelCAD to retopologize the mesh into a cast-ready model.

What jewellery styles does Facetra support?

Traditional Indian styles: Jadau, Polki, Kundan, Meenakari, Temple, Rajwadi. Modern styles: solitaire, halo, cluster, tennis, cocktail. The rendering vocabulary understands craft terms like chaton, bezel, prong, jaali, nakashi and meena work.

What 3D formats does Facetra export?

GLB (default), STL, OBJ, FBX, USDZ, GLTF and 3MF. Two quality tiers: Standard (25 credits, 4K PBR) and HD (35 credits, native 8K PBR).

How much does Facetra cost?

Credit-based pricing. Studio render 1 credit, multi-view sheet 4 credits, CAD-ready 1 credit, 3D mesh 25-35 credits. Credit packs: Starter ₹1,500 for 100 credits (₹15/credit), Studio ₹7,000 for 500 credits (₹14/credit), Atelier ₹15,600 for 1,300 credits (₹12/credit). New signups get 10 free credits.

Is my design confidential on Facetra?

Yes. Every design lives in a private per-user storage bucket. Nothing is public, nothing is shared with other users, nothing is used to train shared AI models.

Workflow

How to Prepare Jewellery Images for E-commerce — Specs, Sizes, Backgrounds

Published 25 February 2026 7 min read

TL;DR

Shopify, Amazon, Flipkart, Meesho, Nykaa Fashion and Ajio — each marketplace has its own image specs (aspect ratio, background, DPI). A cheat-sheet + the two-step AI pipeline that produces marketplace-ready images from any product photo.

Marketplace-specific image specs

E-commerce doesn't accept "close enough". Every marketplace has hard-coded image specs — miss them and your listing either gets rejected outright or looks like a broken tile next to competitor SKUs. Here's the current cheat-sheet (verify before publishing; marketplaces change specs quietly).

Platform	Aspect / Size	Background	Notes
Amazon India	1000×1000 minimum, square	Pure white RGB 255,255,255	Piece fills 85% of frame
Flipkart	500×500 minimum, square	White	2000×2000 recommended for zoom
Meesho	4:5 or 1:1	White	Model images allowed on secondary slots
Nykaa Fashion	1080×1350 (4:5) hero	White or lifestyle	Editorial secondary allowed
Ajio	1080×1440	White or grey	Lifestyle carousel
Shopify (own store)	2048×2048	Your choice	Retina + zoom-friendly
Etsy	2000×1600 (5:4)	Your choice	Story-driven allowed
Instagram Shop	1080×1080	Your choice	Consistent grid look
Google Merchant	800×800 minimum	White preferred	Solid backgrounds only

The 2-step AI pipeline

Instead of Photoshopping 8 variants per SKU, do this:

Photo edit once — upload the raw shot to Facetra Photo Editing, run background = white, lighting correction, colour = correct karat pipeline, cleanup. Export at Ultra-HD (2048px) as your master.
Sizing pass — feed the master to Custom Sizing and get every marketplace variant in one click. The piece's centre of visual weight is preserved; framing is recomposed per platform.

Common rejection reasons + how to avoid them

Off-white background — Amazon rejects if pixels are not pure 255,255,255. Facetra's white-catalog style outputs exact-white by default.
Piece too small in frame — most platforms want 80-90% frame fill. Custom Sizing keeps the piece at your chosen fill ratio across every variant.
Wrong aspect ratio — Meesho and Nykaa are 4:5, not square. Cropping a square image loses composition. Regenerate from the master.
Watermarks — Amazon rejects any watermark, logo or promotional text on the hero image. Keep those for secondary slots.
Wrong JPEG quality — export at 85-95 for the balance of file size and detail. Facetra defaults to this.