GEO visibility ranking
Dénia restaurants most mentioned by AI
This ranking orders Dénia restaurants by combining three factors: how many times the AI assistants analyzed mention them, at what position within the response, and with what tone. It doesn't measure food or service quality, and it isn't a review: it measures visibility and recommendation frequency under this experiment's conditions.
Top 3
This cut's podium
The three restaurants with the highest score in this data cut. The score combines mentions, average position, and tone, explained further down this page.
Quique Dacosta
54 mentions · average position 2.8
Restaurante Mena
77 mentions · average position 3.25
El Pegolí
51 mentions · average position 3.43
Summary
Restaurante Mena tops the ranking with 77 mentions across the assistants analyzed, followed by Quique Dacosta and El Pegolí. This cut's full dataset crosses 451 queries and extracts 1,278 mentions across 498 distinct restaurants, though only 236 of those queries returned capturable text (see the data quality notice below before comparing figures across groups).
How it's calculated
How a restaurant's ranking position is calculated
A restaurant's position no longer depends only on how many times it's mentioned. We calculate it by combining three factors into a single score, so mention count alone can't decide the result.
01
Number of mentions
The base of the score: how many times an assistant recommended that restaurant across this cut's 451 queries. More mentions means a higher starting score.
02
Average position
Appearing among the earliest recommendations in a response adds up to 30% on top of the base score; consistently appearing near the end subtracts up to 30%. We compare each restaurant's average position against the average position across every mention in this cut, not a fixed value.
03
Tone
If most of a restaurant's mentions are positive, it adds up to 30% more; if most are negative, it subtracts up to 30%. Neutral mentions don't move the score either way.
In one line
Score = mentions × position bonus (up to ±30%) × tone bonus (up to ±30%). Because these multiply rather than add, a restaurant with very few mentions can't outrank one with many mentions just by having a great average position or tone: both factors adjust the score, they don't replace it.
All 498 restaurants mentioned in this cut, with their score and fixed ranking position.
Click any column header to sort the table by that field. The "#" column always shows each restaurant's fixed position in the overall ranking, even when the table is sorted by another field.
| 1 | Restaurante Mena | 77 | 3.25 | 66+ 11= 0- | DeepSeek (API), OpenAI, with web search | 8 | 102.6 |
| 2 | Quique Dacosta | 54 | 2.8 | 35+ 13= 6- | DeepSeek (API), OpenAI, with web search | 7 | 68.5 |
| 3 | El Pegolí | 51 | 3.43 | 48+ 3= 0- | DeepSeek (API), OpenAI, with web search | 7 | 68.4 |
| 4 | Casa Federico | 52 | 4.52 | 48+ 4= 0- | DeepSeek (API), OpenAI, with web search | 8 | 64.1 |
| 5 | La Seu | 49 | 4.1 | 44+ 5= 0- | DeepSeek (API) | 7 | 62.0 |
| 6 | El Baret de Miquel Ruiz | 36 | 3.69 | 32+ 4= 0- | DeepSeek (API), OpenAI, with web search | 6 | 46.8 |
| 7 | El Raset | 32 | 4.56 | 24+ 8= 0- | DeepSeek (API), OpenAI, with web search | 6 | 37.8 |
| 8 | Casa Pepa | 21 | 2.52 | 18+ 3= 0- | DeepSeek (API) | 4 | 29.4 |
| 9 | Kokoro | 17 | 2.59 | 14+ 3= 0- | DeepSeek (API) | 2 | 23.5 |
| 10 | Sakura Sushi | 16 | 3.13 | 13+ 3= 0- | DeepSeek (API) | 2 | 21.3 |
| 11 | Il Forno | 13 | 2.31 | 13+ 0= 0- | DeepSeek (API) | 1 | 19.1 |
| 12 | Tasca Eulalia | 13 | 4.54 | 12+ 1= 0- | DeepSeek (API), OpenAI, without web search, OpenAI, with web search | 4 | 16.0 |
| 13 | El Marino | 11 | 5.27 | 11+ 0= 0- | DeepSeek (API), OpenAI, with web search | 4 | 13.0 |
| 14 | Taco Loco | 9 | 2.44 | 5+ 4= 0- | DeepSeek (API) | 1 | 11.8 |
| 15 | La Cuina de Pepa | 11 | 5.82 | 8+ 3= 0- | DeepSeek (API) | 4 | 11.7 |
| 16 | La Cantina de Denia | 8 | 2.75 | 8+ 0= 0- | DeepSeek (API) | 3 | 11.4 |
| 17 | Restaurante Chino Gran Muralla | 8 | 2.38 | 7+ 1= 0- | DeepSeek (API) | 1 | 11.3 |
| 18 | Restaurante Nou Romá | 9 | 4.33 | 7+ 2= 0- | DeepSeek (API), OpenAI, with web search | 4 | 10.9 |
| 19 | Ca Nano | 10 | 6.1 | 9+ 1= 0- | DeepSeek (API) | 4 | 10.8 |
| 20 | Aticcook | 8 | 3.25 | 7+ 0= 1- | DeepSeek (API), OpenAI, with web search | 3 | 10.4 |
| 21 | Casa Benito | 7 | 2.14 | 7+ 0= 0- | DeepSeek (API) | 4 | 10.4 |
| 22 | Mercat Municipal de Dénia | 10 | 5.9 | 6+ 4= 0- | DeepSeek (API) | 4 | 10.2 |
| 23 | Peix & Brases | 7 | 2.43 | 6+ 1= 0- | OpenAI, with web search | 4 | 9.9 |
| 24 | Pizzería La Trattoria | 6 | 1 | 6+ 0= 0- | DeepSeek (API) | 1 | 9.6 |
| 25 | MESA La Marina | 6 | 1.33 | 6+ 0= 0- | OpenAI, with web search | 2 | 9.4 |
| 26 | La Barraca | 8 | 4.88 | 6+ 2= 0- | DeepSeek (API) | 4 | 9.2 |
| 27 | Restaurante Sendra | 7 | 5 | 7+ 0= 0- | DeepSeek (API) | 4 | 8.5 |
| 28 | Kabuki | 6 | 2.83 | 5+ 1= 0- | DeepSeek (API) | 2 | 8.2 |
| 29 | Meshi.co | 6 | 2.67 | 4+ 2= 0- | OpenAI, with web search | 3 | 7.9 |
| 30 | XqNo? Mexiterránea | 5 | 1.2 | 5+ 0= 0- | OpenAI, with web search | 1 | 7.9 |
| 31 | Mishima | 6 | 3 | 4+ 2= 0- | DeepSeek (API) | 2 | 7.8 |
| 32 | Ca Pepa Teresa | 7 | 6 | 5+ 2= 0- | DeepSeek (API), OpenAI, without web search, OpenAI, with web search | 4 | 7.3 |
| 33 | BonAmb | 6 | 4.67 | 5+ 1= 0- | DeepSeek (API) | 1 | 7.2 |
| 34 | Restaurante Noguera | 6 | 5.5 | 6+ 0= 0- | DeepSeek (API) | 3 | 7.0 |
| 35 | La Trattoria | 5 | 3 | 4+ 1= 0- | DeepSeek (API) | 2 | 6.7 |
| 36 | La Tasca del Puerto | 6 | 5.17 | 4+ 2= 0- | DeepSeek (API) | 2 | 6.6 |
| 37 | Wok Denia | 5 | 3.2 | 4+ 1= 0- | DeepSeek (API) | 2 | 6.6 |
| 38 | Sushi Dénia | 5 | 1.8 | 2+ 3= 0- | DeepSeek (API) | 1 | 6.5 |
| 39 | El Poblet | 6 | 4.33 | 2+ 4= 0- | DeepSeek (API) | 2 | 6.5 |
| 40 | Agua de Mar | 6 | 5.5 | 4+ 2= 0- | DeepSeek (API), OpenAI, with web search | 3 | 6.4 |
| 41 | Umai | 5 | 4 | 4+ 1= 0- | DeepSeek (API) | 1 | 6.2 |
| 42 | Casa Mexicana | 5 | 3.6 | 3+ 2= 0- | DeepSeek (API) | 1 | 6.1 |
| 43 | Casa Papat | 4 | 2.25 | 4+ 0= 0- | OpenAI, with web search | 2 | 5.9 |
| 44 | La Trattoria di Denia | 4 | 2.25 | 4+ 0= 0- | DeepSeek (API) | 1 | 5.9 |
| 45 | La Cantina Mexicana | 4 | 1.5 | 3+ 1= 0- | DeepSeek (API) | 1 | 5.8 |
| 46 | JapiJapo | 4 | 2.5 | 4+ 0= 0- | OpenAI, with web search | 2 | 5.8 |
| 47 | Casa Benjamín | 4 | 2.75 | 4+ 0= 0- | DeepSeek (API) | 4 | 5.7 |
| 48 | NOMADA Dénia | 4 | 2 | 3+ 1= 0- | OpenAI, with web search | 1 | 5.7 |
| 49 | Mizuki | 4 | 3 | 4+ 0= 0- | DeepSeek (API) | 2 | 5.6 |
| 50 | Restaurante Balandros | 4 | 3 | 4+ 0= 0- | OpenAI, with web search | 3 | 5.6 |
Showing 1–50 of 498 restaurants.
Page 1 of 10
"Average position" is the average order in which the restaurant appears within the responses that mention it (1 = first mention in the response), not a search-engine ranking position. "Cuisines" counts how many of the experiment's nine query categories it was mentioned under.
Frequently asked questions
Before drawing conclusions from the ranking
Is the top restaurant the best restaurant in Dénia?
Not necessarily. It's the restaurant the AI assistants analyzed cited most often under this experiment's conditions. It may or may not match actual quality: this ranking doesn't measure that.
Can a restaurant with fewer mentions rank above one with more?
Yes. The score combines mentions, average position, and tone (see "How a restaurant's ranking position is calculated" above), so a restaurant with fewer mentions but a better average position and mostly positive tone can outrank one with more mentions. Mention count is still the heaviest factor: the other two adjust the score by up to 30% each, they don't flip it.
My restaurant isn't listed, what should I do?
It may not have been mentioned in any of this cut's 451 queries, or it came up in one of the 215 calls with no capturable text (see the data quality notice). The table below includes all 498 mentioned restaurants: sort it by name to find yours faster, or download the full dataset and search it there. If it's still missing, write to us through the contact form and we'll look into it.
Data quality notice
Of the 451 calls in this cut, 236 returned capturable text. The gap isn't spread randomly: it's concentrated in three specific day, time-slot, and assistant combinations (detail below) totaling 217 calls, of which only 2 returned text. Those calls completed their run and used tokens, but the text couldn't be recovered due to a capture-pipeline gap, not because the assistant recommended nothing. Every figure on this page counts them toward the total queries, but they can't contribute mentions.
- Friday · Afternoon · DeepSeek (API): 129 of 129 calls with no capturable text
- Saturday · Morning · OpenAI, without web search: 43 of 44 calls with no capturable text
- Sunday · Midday · OpenAI, without web search: 43 of 44 calls with no capturable text
Check how this was calculated
Every figure in this ranking comes from a documented process: what was asked, how the mention was extracted, and what limitations it has. It's all in the methodology.