The Citation DeskIndex No. 02 · August 2026
We ran 300 real buying questions across five AI engines and recorded every jewellery brand that came up. A handful of names dominate — and one online-first brand outranks almost every legacy house. Here is exactly who AI names, and who it leaves out.
India's jewellery market is growing at an estimated 12–14% a year and is projected to reach $130–140 billion by 2029 — and the fastest-shifting part is the move online. The organised, branded share is expected to climb from 37% in 2024 to 43–47% by 2029, growing at roughly 16–18% a year — faster than the market overall.1 Buyers increasingly discover and shortlist designs online before they visit a store,2 and a fast-rising share now ask an AI assistant — "best online jewellery brand", "where to buy a lab-grown diamond" — and act on the shortlist it gives back. As one analyst put it, legacy brands increasingly face competition from agile D2C players.3
Sources: 1. Redseer Research & Analysis, cited in BlueStone's regulatory filing (RHP), 2025. 2. IMARC Group, India Jewellery Market analysis, 2026. 3. Euromonitor International, Jewellery in India, 2025.
Which means one question increasingly decides whether a brand makes the buyer's shortlist: when someone asks AI, does your brand get named — or does your competitor? This index measures exactly that.
Every brand's score is the share of the 300 answers it was named in, counted once per answer. The top 20 of 289 brands that surfaced:
289 distinct brands appeared in total. Below the top tier, visibility falls away fast — most brands appeared in under 7% of answers.
GIVA (50%) sits second overall — ahead of Tanishq, BlueStone, Kalyan and Malabar. A digital-first brand founded in the last decade is named by AI more often than century-old jewellery houses. It's the clearest proof in the data that AI visibility is earned online, not inherited from legacy or store count. CaratLane (64%) leads overall; Palmonas (25%) is another young online-first brand punching far above its age.
The five engines don't return the same brands. BlueStone is named 30 times on Perplexity but only 9 times on Gemini. Mia by Tanishq drops from 27 on ChatGPT to 9 on Gemini. Kalyan and Malabar are strong on Claude (27 each) but weak on Google (7 and 10). A single "AI visibility" number would hide this entirely — which is why every brand needs a per-engine picture, not one score.
Segment by segment, different brands surface. Within this index's lab-grown-diamond prompts, Limelight and Fiona come up most — not CaratLane. Across the silver prompts, GIVA appears in nearly every answer. Across the gold prompts, Tanishq comes up most often. On problem-first questions like "how do I know if a brand is genuine," AI often names no brand at all. Each segment behaves differently — and each is uncontested space a focused brand can still win.
Mention counts for the top 10 brands, out of 60 answers per engine. Green = that brand's strongest engine; red = its weakest.
| Brand | ChatGPT | Perplexity | Gemini | Claude | Google AI |
|---|---|---|---|---|---|
| CaratLane | 39 | 41 | 33 | 45 | 35 |
| GIVA | 26 | 33 | 26 | 36 | 29 |
| Tanishq | 29 | 31 | 25 | 36 | 23 |
| BlueStone | 28 | 30 | 9 | 28 | 29 |
| Mia by Tanishq | 27 | 24 | 9 | 19 | 20 |
| Kalyan Jewellers | 19 | 20 | 12 | 27 | 7 |
| Malabar Gold & Diamonds | 16 | 16 | 8 | 27 | 10 |
| Palmonas | 12 | 15 | 10 | 21 | 17 |
| Limelight Diamonds | 8 | 7 | 8 | 13 | 11 |
| Voylla | 5 | 15 | 5 | 17 | 1 |
The spread across a single row is the story: strong on one engine, near-invisible on another. Fixing that gap is brand-specific work.
The overall leaderboard hides how differently each segment behaves. These are signals from this general index's segment-specific prompts — directional, not definitive rankings (each segment deserves its own dedicated index). A few of the clearest:
Within this index's lab-grown prompts, Limelight Diamonds, Fiona Diamonds, Aukera and Jewelbox came up most — the legacy houses barely featured. A signal that the fastest-growing segment is being led by specialists, and a case for a dedicated lab-grown index.
Across this index's silver prompts, GIVA appeared in every answer, with Shaya and Mia by Tanishq close behind. Online-first brands show up strongest in the everyday and silver space.
Across the gold and natural-diamond prompts, Tanishq came up most, with CaratLane and the legacy houses behind. Trust-heavy, high-value segments still lean toward the established names.
We wrote 60 real buyer questions spanning the whole journey — category discovery, feature and type, lab-grown and natural diamond, comparisons, occasions, problem-first and pricing — and ran each across ChatGPT, Perplexity, Gemini, Claude and Google AI Overview in fresh, logged-out sessions. Every brand named in every answer was recorded and scored one answer at a time, in order. This is not an automated scan — each of the 300 answers was read and judged one at a time. That's what makes it accurate.
This index is the category-wide view. A full audit shows your brand: where you stand on each engine, who's named instead of you, the questions you're missing from, and a prioritized plan to get recommended more often. Read and judged one answer at a time — which is why only a few run at once.
Request your audit →