Real estate brokerages AI recommends, by metro and brand

The GeoGenius Brokerage Rankings roll the AI Citation Index up to core brokerage brands: in each of 49 US metros, the real estate agents that ChatGPT, Claude, Gemini and Grok name are grouped by the brokerage they work for. Across 1588 named agents, Keller Williams has the most (160 across 44 metros), followed by Compass (140 in 28) and Sotheby's International Realty (88 in 30). The rankings measure presence in AI recommendations — how many of a brokerage's agents the engines actually name — not sales volume, headcount or quality, and payment plays no part. Each metro page ranks every brokerage the engines named there; each brand page shows where it leads.

49 metros · 439 brokerages · 1588 named agents · 204 without a brokerage on record · updated Sep 4, 2026 · stats.json · Agent index

Brokerages with the most AI-named agents

Top 25 of 439 brokerages across 49 metros
RankBrokerageAgents namedMetrosLeads inResponsesEngines
1Keller Williams national 1604411593ChatGPT Claude Gemini Grok
2Compass national 140289531ChatGPT Claude Gemini Grok
3Sotheby's International Realty national 88302354ChatGPT Claude Gemini Grok
4Coldwell Banker national 83266352ChatGPT Claude Gemini Grok
5RE/MAX national 70281292ChatGPT Claude Gemini Grok
6Windermere national 69105319ChatGPT Claude Gemini Grok
7eXp Realty national 48272190ChatGPT Claude Gemini Grok
8Redfin national 3116297ChatGPT Claude Gemini Grok
9Berkshire Hathaway HomeServices national 26151100ChatGPT Claude Gemini Grok
10The Agency national 189192ChatGPT Claude Gemini Grok
11Real Broker national 1810075ChatGPT Claude Gemini Grok
12John L. Scott national 188071ChatGPT Claude Gemini Grok
13Christie's International Real Estate national 137159ChatGPT Claude Gemini Grok
14Douglas Elliman national 113041ChatGPT Claude Gemini Grok
15Corcoran national 115035ChatGPT Claude Gemini Grok
16Levin Rinke Realty independent 91137ChatGPT Gemini Grok
17Century 21 national 96122ChatGPT Claude Gemini Grok
18Every Door Real Estate independent 85047ChatGPT Claude Gemini Grok
19ERA Real Estate national 73131ChatGPT Gemini Grok
20@properties national 72027ChatGPT Claude Gemini Grok
21Real Residential independent 63041Claude Gemini Grok
22Engel & Völkers national 55018ChatGPT Claude Gemini Grok
23ReeceNichols national 51018Gemini Grok
24The Brokerage House independent 51116Gemini Grok
25Edina Realty national 51112Gemini

414 more brokerages with fewer named agents.

Browse by state

  • Alabama — 1 metro · 16 named agents · most: RealtySouth (4)
  • Arizona — 1 metro · 30 named agents · most: eXp Realty (7)
  • California — 3 metros · 97 named agents · most: Compass (12)
  • Colorado — 1 metro · 34 named agents · most: Compass (5)
  • Connecticut — 1 metro · 24 named agents · most: Coldwell Banker (6)
  • Florida — 3 metros · 81 named agents · most: Keller Williams (13)
  • Georgia — 2 metros · 35 named agents · most: Keller Williams (6)
  • Idaho — 1 metro · 34 named agents · most: Keller Williams (8)
  • Illinois — 1 metro · 37 named agents · most: Compass (10)
  • Indiana — 1 metro · 19 named agents · most: Century 21 (3)
  • Iowa — 1 metro · 34 named agents · most: RE/MAX (14)
  • Louisiana — 1 metro · 25 named agents · most: Keller Williams (3)
  • Maine — 1 metro · 26 named agents · most: Redfin (4)
  • Massachusetts — 1 metro · 23 named agents · most: Sotheby's International Realty (4)
  • Michigan — 1 metro · 36 named agents · most: eXp Realty (5)
  • Minnesota — 1 metro · 30 named agents · most: Edina Realty (5)
  • Missouri — 1 metro · 30 named agents · most: Keller Williams (12)
  • Montana — 1 metro · 17 named agents · most: Christie's International Real Estate (4)
  • New Jersey — 1 metro · 14 named agents · most: Coldwell Banker (5)
  • New Mexico — 1 metro · 24 named agents · most: Sotheby's International Realty (11)
  • New York — 1 metro · 44 named agents · most: Compass (10)
  • North Carolina — 1 metro · 31 named agents · most: Keller Williams (3)
  • North Dakota — 1 metro · 25 named agents · most: Berkshire Hathaway HomeServices (4)
  • Ohio — 1 metro · 31 named agents · most: The Brokerage House (5)
  • Oregon — 1 metro · 27 named agents · most: Redfin (6)
  • Pennsylvania — 3 metros · 92 named agents · most: Coldwell Banker (9)
  • South Dakota — 1 metro · 25 named agents · most: Keller Williams (7)
  • Tennessee — 1 metro · 50 named agents · most: Compass (15)
  • Texas — 3 metros · 118 named agents · most: Keller Williams (16)
  • Vermont — 1 metro · 23 named agents · most: Coldwell Banker (5)
  • Virginia — 1 metro · 18 named agents · most: Long & Foster (4)
  • Washington — 9 metros · 438 named agents · most: Windermere (68)

Browse by metro

19 more metros are listed by state above.

How this index is built

GeoGenius asks ChatGPT, Claude, Gemini and Grok the questions a buyer or seller would ask about a market and records every agent each engine names. An agent is published once at least two responses, or two different engines, have named them; their brokerage, as the engines state it, is resolved to its core brand. A metro is ranked once at least 5 named agents span 3 or more brokerages; 127 more metros are recorded but not yet ranked. The method mirrors our published 36-market study at https://geogenius.ai/ai-report.

Frequently asked questions

What are the Brokerage Rankings?
A brand-level view of the GeoGenius AI Citation Index. For every covered metro, the agents that ChatGPT, Claude, Gemini and Grok name are grouped by the brokerage they work for, rolled up to the core brand, and ranked by how many named agents each brand has.
How are brokerages ranked?
By named agents — agents the engines named in at least two responses, or by two different engines — then by distinct responses, then by distinct engines. Offices, teams and franchise affiliates count toward their parent brand; independents rank under their own name. Payment and advertising play no part.
Does this measure sales volume or brokerage size?
No. A large brokerage with few AI-named agents ranks below a small one whose agents the engines recommend. It measures presence in AI recommendations only.
Can I use this data?
Yes. Each metro has a JSON endpoint at /brokerages/{metro}.json with the ranked brands (name, kind, named agents, share, responses, engines) and /brokerages/stats.json has the totals. Contact details are never published.

Get more of your roster on these lists