The Wine AI Visibility Index 2026 — Top 25 Wine Brands by AI Citation Share

Last updated: 2026-09-04

Domaine de la Romanée-Conti had the highest measured appearance rate in this wine-brand study: 36.9% of 268 answers returned across four engines. Its 95% interval was 31.4% to 42.9%. The result places ultra-luxury and investment-oriented names ahead of mass-market wine brands within these measured consumer-intent questions.

The gap between brand mention and brand-site citation is large. Wine-Searcher had a 19% owned-domain citation rate among answers with retrieval evidence, while Domaine de la Romanée-Conti had a 0% owned-domain citation rate despite leading appearance. A wine brand can be named often without its own website supplying the cited evidence.

For category teams, the risk is clear: an answer may steer selection toward a competitor or a critic and comparison site without citing the brand's own material.

The study measured 24 prompts three times in Perplexity (sonar-pro), ChatGPT (OpenAI), Gemini (Google), and Claude (Anthropic). It returned 268 answers from 288 attempted runs.

Which wine brands lead AI answers?

Domaine de la Romanée-Conti ranked first, followed by Château Lafite Rothschild at 27.2% and Château Margaux at 23.5%. Penfolds ranked fourth at 21.6%, and Screaming Eagle ranked fifth at 17.9%. These are appearance rates calculated from answers returned, not citation rates.

The ranking measures whether an answer named a brand. It does not measure sales, consumer preference, critic scores, or category market share. It describes only the 24 measured prompts on September 4, 2026.

Wine brand appearance leaderboard

Appearance rate is the share of the 268 returned AI answers in which each brand was named.

The top brands are not always statistically separable in this sample. Domaine de la Romanée-Conti's 31.4% to 42.9% interval does not overlap Château Lafite Rothschild's 22.3% to 32.9% interval, but Château Lafite Rothschild and Château Margaux overlap. Château Margaux, Penfolds, and Screaming Eagle also have overlapping intervals with adjacent leaders. Those differences are within the margin of this sample.

The leading wine brands show different engine patterns even when each appears across all four systems. The profiles below use answers returned for appearance rates. Owned-domain citation rates use answers with retrieval evidence.

Where do the engines disagree on wine brands?

The engines disagreed most sharply on individual prestige, investment, and comparison entities. Each engine had a similar number of returned answers: Perplexity returned 68, ChatGPT 66, Gemini 67, and Claude 67. The differences below therefore reflect naming patterns, not a large difference in answer volume.

Engine-by-brand appearance pattern

Appearance rate by engine

Each cell uses that engine's own answer count as its denominator, so engines can be compared row by row.

Gemini concentrated several results. Liv-ex appeared in 25 of 67 Gemini answers and none in Perplexity or Claude. Wine-Searcher appeared in 20 Gemini answers, compared with 1 each in Perplexity and ChatGPT and none in Claude. Bodega Catena Zapata appeared in 14 Gemini answers, while ChatGPT and Claude did not name it.

ChatGPT concentrated several Bordeaux and California prestige names. Château Latour appeared in 20 of 66 ChatGPT answers but only 5 Perplexity, 6 Gemini, and 3 Claude answers. Château Mouton Rothschild appeared in 19 ChatGPT answers, with 6, 3, and 5 appearances in Perplexity, Gemini, and Claude. Screaming Eagle appeared in 21 ChatGPT answers, compared with 8, 9, and 10 in the other engines.

Perplexity had its own concentration pattern. Catena appeared in 21 of 68 Perplexity answers, while ChatGPT and Gemini each named it once and Claude named it twice. Opus One appeared in 19 Perplexity answers, compared with 8 in ChatGPT, 1 in Gemini, and 3 in Claude. Vega Sicilia appeared in 14 Perplexity answers and none in Gemini.

Some names had broad engine coverage but uneven rates. Domaine de la Romanée-Conti appeared in all four engines, ranging from 21 Gemini answers to 28 each in Perplexity and ChatGPT. Château Lafite Rothschild appeared in all four, ranging from 11 Gemini answers to 23 Perplexity answers. A gap can reflect a system's model and retrieval tool, not a durable difference in brand visibility outside this study.

Which demand clusters favor each wine brand?

The demand-cluster results identify which brands lead specific groups of measured questions rather than a universal category winner. Prestige, luxury, investment, and cult-wine questions pull leading appearances toward Domaine de la Romanée-Conti, Château Lafite Rothschild, Château Margaux, Penfolds, and Screaming Eagle.

Brand leaders by demand cluster

Leaders by demand cluster

The same answers regrouped by the demand cluster each prompt was sampled for; cluster rates use that cluster's own answer count.

General buyer-intent questions and specific need questions measure different forms of demand. The chart separates those groups. Read the results as prompt-level strengths: a brand can lead one cluster while remaining less visible in another. The category includes value, premium, luxury, ultra-luxury, box wine, premium rosé, premium import wine, premium Chardonnay, cult wine, prestige Champagne, investment-grade wine, low-alcohol wine, and moderation-adjacent wine.

Which sites supply the citations?

Wine-Searcher had the highest owned-domain citation rate among eligible entities: 19% of 268 answers with retrieval evidence. Liv-ex followed at 8.2%, and Penfolds had 3%. These rates use answers with retrieval evidence, not all answers returned.

Owned-domain citation rates

Share of the 268 answers with retrieval evidence that cited a URL on the brand's own website.

Wine-Searcher also had 51 owned-domain citations and 1.8% of all citations. Liv-ex had 22 owned-domain citations and 0.8% of all citations. Among leading prestige brands, owned-site citation remained low: Château Lafite Rothschild and Château Margaux each had 4 owned-domain citations, while Domaine de la Romanée-Conti and Screaming Eagle had none.

Publisher citations leaned toward wine media, investment sites, lifestyle publications, reference pages, retailers, and consumer discussion. Wine Enthusiast appeared in 96 answers with citations, or 35.8% of answers with retrieval evidence. Vinovest appeared in 80, VinePair in 68, Forbes in 57, and Wine-Searcher in 51.

Most-cited publisher domains

Most-cited publishers

Publishers cited as evidence, counted once per answer rather than once per link.

The source list shows where systems found support in this run. It does not measure editorial quality, commercial influence, or traffic. It does show that brand-owned sites compete with established third-party material for citation space.

What does this mean for wine brands?

Interpretation: prestige recognition and source ownership are separate measures. Domaine de la Romanée-Conti led appearance but had no owned-domain citations. Wine-Searcher had a lower appearance rate than the top wine brands but the highest owned-domain citation rate among eligible entities.

Interpretation: category teams need separate checks for whether a brand is named and whether its own product, provenance, critic-score, food-pairing, gifting, scarcity, and occasion pages are cited. A strong appearance result does not establish that a brand controls the supporting information.

Interpretation: the sharp engine splits warrant engine-specific review. Liv-ex, Wine-Searcher, Catena, Opus One, Château Latour, and Screaming Eagle each had an appearance pattern concentrated in one system. A category team that checks only one engine may miss a material blind spot in another.

How was this wine-brand study measured?

This flagship-depth study tested 24 stratified prompts three times per engine across Perplexity (sonar-pro), ChatGPT (OpenAI), Gemini (Google), and Claude (Anthropic). The study attempted 288 answers and returned 268. Appearance rates use answers returned as their denominator. Citation and source rates use the 268 answers with retrieval evidence as their denominator.

Measurement summary

How this study was measured

• Sampling frame fixed first: 24 prompts allocated across the demand clusters before anything was asked.
• Same questions put to every engine: Perplexity (sonar-pro), ChatGPT (OpenAI), Gemini (Google), Claude (Anthropic), 3 trial(s) each.
• 268 answers returned of 288 attempted, 268 with retrieval evidence.
• Product lines rolled up to parent brands using a registry fixed before scoring.
• Appearance rate reported with a 95% Wilson interval; citation metrics counted only on answers with evidence.

The study measures named brands and cited domains in this fixed prompt set on September 4, 2026. It does not measure all wine searches, consumer demand, sales, product quality, or future performance. Brand names were normalized, including variants such as Penfolds Grange under Penfolds and Catena Zapata under Catena.

Charts and data tables

These are not brands in the category, so they are measured the same way but kept out of the brand leaderboard.

Methodology appendix

This appendix is generated directly from the measurement record for this study. It is included so any brand named here can reproduce or challenge the result.

What was measured

• Category: wine brands
• Measurement date: 2026-09-04
• Depth: flagship — 24 prompts, 3 trial(s) per prompt per engine, 4 engines
• Answers attempted: 288
• Answers returned (appearance denominator): 268
• Answers with retrieval evidence (citation denominator): 268 (93.1% of attempted)
• Answers that failed outright: 20
• Analysis version: study-methodology-v2 · entity registry: entity-registry-2026-08-27

Claim limit: Flagship depth: 24 stratified prompts measured three times per engine, so both between-prompt and repeat-generation variation are observable.

How the prompts were sampled

Fixed sampling frame: 13 of 13 client-named subcategories are allocated a guaranteed cell each, then a documented set of generic commercial demand clusters (general best, primary use case, specific need, quality tier, price/value, beginner, comparison, durability, specialist, emerging, retail channel, trust) receives the remaining allocation round-robin. The prompt writer supplies wording inside these cells only and does not choose the category's demand structure.

Engines and systems measured

Results are not strictly comparable across engines: each engine uses a different model, a different retrieval tool and a different answer format, so a gap between two engines can reflect system design as much as brand visibility.

How each metric is defined

• Appearance rate — the share of returned answers that named the brand. Measured on every answer the engine returned, whether or not it searched the web, so an engine's retrieval behaviour does not silently change the denominator.
• Macro appearance rate — the mean of the per-prompt appearance rates, so each prompt counts once regardless of how many engines answered it.
• 95% interval — Wilson score interval on the appearance rate. Where two brands' intervals overlap, this study does not support a claim that one is more visible than the other.
• Owned-domain citation rate — the share of answers with retrieval evidence that cited a URL on the brand's own domain. This is a narrow measure of the brand's own site being used as a source; it is not a measure of total mentions.
• Share of all citations — the brand's own domain as a share of all attributed publisher citations in the study.
• Not measured — the brand's official domain could not be established, so no citation result exists. This is never reported as 0%.
• 0% — the metric could be evaluated and the event did not occur.
• Average position — mean position of the brand's first mention in the answer's ordered list of named brands.
• Cross-engine consistency — the share of engines that named the brand at least once.
• Repeat consistency — the share of repeated prompt/engine cells where the brand appeared in every trial. Lower values mean the engine's answer is unstable rather than the brand being absent.

Evidence rules

• An answer counts as having retrieval evidence only when the engine returned a retrieval receipt (search annotations or grounding records). A URL the model merely typed into its prose is not treated as evidence that it searched.
• Sources are attributed to the publisher behind each retrieval receipt. Intermediary and redirect hosts are never reported as publishers.
• Publisher identity could not be established for 0 retrieval receipt(s); those are disclosed here rather than attributed or dropped.
• Source counts are answer-level: a publisher counts once per answer in which it was cited, not once per URL.

Entity handling

Product lines and model names roll up to their parent brand for brand-level metrics, using a registry fixed before scoring. Nothing is merged on the basis of this study's own counts alone.

No roll-ups were applied in this study.

Held for human classification before the next study: Yellow Tail, Clos du Bois Wines, Beringer Vineyards, Barefoot Cellars, Cupcake Vineyards, Apothic Wines, Kendall-Jackson, J. Lohr Vineyards & Wines, Costco's Kirkland Signature, Domaine de la Romanée-Conti, Château Lafite Rothschild, Château Margaux, Vega Sicilia, Opus One, Penfolds, Moët & Chandon, Château Latour, Screaming Eagle, Harlan Estate, Bodegas Vega Sicilia, Opus One Winery, Château Mouton Rothschild, Marchesi Antinori, Catena Zapata, Krug, Domaine Leflaive, Joseph Drouhin, Famila Torres, Status, Notable Label, Liv-ex, Wine-Searcher, Vin-X Fine Wine Investment, Dom Pérignon, Veuve Clicquot, Bollinger, Cristal, Armand de Brignac, Prestige Tier, Decanter & Wine Advocate.

Limits of these findings

• This is a point-in-time measurement of what the named systems returned for these exact prompts on 2026-09-04. Engine behaviour changes without notice.
• The prompt set is a stratified sample of category demand, not a census. Results describe these prompts and should not be read as representative of all searches in the category.
• Where intervals overlap, differences between brands are not statistically distinguishable at this sample size.
• A brand not listed here was not measured as absent — it simply was not named in these answers.
• Appearance is a visibility measure, not an endorsement, a quality judgement or a market-share estimate.
• 6 of 49 ranked brand(s) had no official domain established before scoring, so they carry appearance results only and are excluded from citation metrics and source tables rather than being reported as zero.