ChatGPT search research · Comparison

Free vs Plus

The ChatGPT app without a subscription against the paid plan, on the same prompts.

Wave 2026-11Free: gpt-6Plus: gpt-6-thinking100 prompts on bothNew York, US
Answer brands the first search did not name
83%Free
47%Plus
They came after the first search, so the web had a chance to put them there.
Search steps per answer
1.0Free
1.2Plus
About the same depth of search.
First search already names a brand
32%Free
78%Plus
30 of 94 against 78 of 100 answers: candidates chosen before searching.
Answers that opened a page
0%Free
15%Plus
The Free stream shows no page reads.
Brands they agree on
49%of brands, same prompt
For the same prompt: 348 brands named by both, 206 by Free only, 173 by Plus only.
  • The first search already names a brand in 30 of 94 Free answers and 78 of 100 Plus answers.
  • Of the brands in the answer, 83% on Free and 47% on Plus were not in its first search, so search results had a chance to put them there. The rest were in its first search, named before any results.
  • "Best …" articles make up 69% of Free's citations and 33% of Plus's; brands' own sites 11% and 36%.
  • For the same prompt they agree on 49% of the brands they name between them.
FreePlus

Every prompt ends by asking ChatGPT to list the searches it ran. The app hides the text of some searches, mostly the second-layer fan-outs it runs while reasoning, so for the app's query numbers this self-report is the default: it is used when it contains every query seen in the stream (Free 93 of 100, Plus 100 of 100 answers), otherwise only what was seen counts. Where the self-report changes a number, the number without it is shown under it. Search steps, page reads, citations and brands always come from the stream.

How much it searches

Counted from search events, which both surfaces show even when the app hides the query text.

Seen: events in the app's stream or calls in the API response.

Search steps per answer
A step is one search call; it can send several queries at once.
Same prompt: Free higher 3, Plus 27, equal 70
Free 1.0median 1
Plus 1.2median 1
Answers with two or more steps
ChatGPT searched, read the results and searched again.
Free 4%4 of 100
Plus 24%24 of 100
Answers that searched at all
Free 94%94 of 100
Plus 100%100 of 100
Steps with the query text hidden
The step is in the stream, its query text is not.
Free 5%5 of 98 steps
Plus 19%24 of 124 steps

The first search

Does ChatGPT start from brands it already has in mind, or from an open search? The first search step decides it.

A step can hold several queries. It counts as naming a brand when at least one of them does. The step is taken from the stream on every surface. Round 1 of the app's self-report fills in only a first step whose text was hidden: that round also lists product lookups by name that ran in no visible step, so it would overstate how often the first search names a brand.

Names a brand or product
At least one query in the first step names one: the candidates were chosen before searching.
Free 32%30 of 94
Plus 78%78 of 100
Names no brand
A "best …" or topic search, open to whatever comes back.
Free 67%63 of 94
Plus 22%22 of 100
Unknown: the app hid what it typed
The stream shows the first search ran but not its query, so whether it named a brand is unknown. These answers are not judged by their second search instead.
Free 1%1 of 94
Plus 0%0 of 100
First-step queries that name a brand
Share of the queries in the first step, not of answers.
Free 27%36 of 132 queries
Plus 63%191 of 305 queries

From memory to the answer

The first search runs before ChatGPT has seen any results, so the brands it names there come from what the model already knows. How many of them survive into the answer, and how much of the answer they make up?

Brands are matched loosely: “RAV4” in a query matches “Toyota RAV4 Hybrid” in the answer. Answers that did not search, or whose first search hid its text, are left out. On the app, a hidden first step is filled from round 1 of the self-report.

Brands named in the first search
Per answer whose first search is known. Before any results, so they come from what the model already knows.
Free 1.1median 0
Plus 3.0median 3
First-search brands that make the answer
Of the brands it searched for first, the share it names in the answer.
Free 84%83 of 99 brands
Plus 85%254 of 300 brands
Answer brands that were in the first search
Of the brands in the answer, the share it already had in mind before searching. The rest came later: from results or later searches.
Free 17%88 of 524 brands
Plus 53%275 of 521 brands
Answers that drop a first-search brand
Among answers whose first search named a brand: at least one of those brands is missing from the answer.
Free 37%11 of 30
Plus 44%34 of 78

Search journeys

A journey is the searches of one answer in order. Two questions sort them: does the first search already name brands, and does ChatGPT search again once it has seen results?

Most common: never searches for a brand on Free (60% of answers), names brands, searches once on Plus (57%).

A second search is counted from the search steps in the stream, or from a later round of the self-report on the app. Brands are matched loosely, as above.

How the journeys go
Share of answers; the rows add up to 100%.
Names brands, searches once
The first search already names the brands and there is no second one: it checks its own picks in one go, then writes.
Free 29%29 of 100
Plus 57%57 of 100
Names brands, then searches again
The first search names brands; after the results come back it searches again.
Free 1%1 of 100
Plus 21%21 of 100
Starts open, names brands later
The first search names no brand. Brands appear only in a later search, after it has seen results.
Free 3%3 of 100 0% · 0 of 100 without the self-report
Plus 2%2 of 100 0% · 0 of 100 without the self-report
Never searches for a brand
No search names a brand: it takes the options from the lists and pages it found.
Free 60%60 of 100 63% · 63 of 100 without the self-report
Plus 20%20 of 100 22% · 22 of 100 without the self-report
First search hidden
The app hid the text of the first search, so how the journey started is unknown.
Free 1%1 of 100
Plus 0%0 of 100
No search
Answered without searching.
Free 6%6 of 100
Plus 0%0 of 100
What the searches after the first one do
Among answers that searched again and showed what they typed (Free 4, Plus 22). One answer can do several of these. Free searched again in fewer than 10 answers, so its shares rest on very few.
Names a brand the first search did not
Picked up from the results it had just seen, or recalled later.
Free 75%3 of 4
Plus 91%20 of 22
Searches a first-search brand again
It goes back to one of its own picks to check it.
Free 0%0 of 4
Plus 82%18 of 22
Looks up facts or prices
Free 50%2 of 4
Plus 73%16 of 22
Checks a brand on its own
Free 25%1 of 4
Plus 77%17 of 22
Checks rules or official sources
Free 50%2 of 4
Plus 18%4 of 22
Asks for reviews or forums
Free 0%0 of 4
Plus 5%1 of 22
Searches for a list or the topic again
Free 0%0 of 4
Plus 18%4 of 22
One prompt, two journeys: “best place to buy tires online”
A prompt where each surface took its most common journey. Up to 3 queries per search are shown; the answer page has all of them.
Free never searches for a brand, as ChatGPT listed its searches
  1. 2 queries · List
    • best online tire retailers tires
    • best places to buy tires online US shipping installation warranty 2026
Every query of this answer →
Plus names brands, searches once, as ChatGPT listed its searches
  1. 4 queries · List with brands, Brand check, Facts
    • Michelin Defender2 225/65R17 Tire
    • Continental TrueContact Tour 54 225/65R17
    • best places to buy tires online Tire Rack Discount Tire SimpleTire Walmart Costco installation shipping 2026
    • and 1 more
Every query of this answer →

Query kinds

The mix of its queries: each slice is the share of its queries of that kind. Each prompt's answer page shows every query with its kind.

Most common: open list search on Free, brand check on Plus.

Kinds come from the classifier's query labels. On the app the mix includes the self-report, which also covers the steps whose text the app hid; the table gives the mix without it. Lookups the self-report puts in its first round but the stream shows in no step count under all searches only.

Every query of every answer.

Free
Topic search: 7% (16 of 225 queries)Open list search: 55% (124 of 225 queries)List search with its own picks: 10% (23 of 225 queries)Brand check: 10% (23 of 225 queries)Fact lookup: 8% (17 of 225 queries)Comparison: 2% (5 of 225 queries)Reviews and forums: 1% (2 of 225 queries)Other: news, official rules, not labeled: 7% (15 of 225 queries) 55% Open listsearch 10% List withbrands 10% Brandcheck 225 queries Topic search: 7% (16 of 225 queries)Open list search: 55% (124 of 225 queries)List search with its own picks: 10% (23 of 225 queries)Brand check: 10% (23 of 225 queries)Fact lookup: 8% (17 of 225 queries)Comparison: 2% (5 of 225 queries)Reviews and forums: 1% (2 of 225 queries)Other: news, official rules, not labeled: 7% (15 of 225 queries) 55% Open listsearch 10% List withbrands 10% Brand check 225 queries
Plus
Topic search: 1% (5 of 537 queries)Open list search: 10% (53 of 537 queries)List search with its own picks: 12% (63 of 537 queries)Brand check: 36% (194 of 537 queries)Fact lookup: 26% (138 of 537 queries)Comparison: 2% (12 of 537 queries)Reviews and forums: 3% (18 of 537 queries)Other: news, official rules, not labeled: 10% (54 of 537 queries) 36% Brandcheck 26% Factlookup 12% List withbrands 537 queries Topic search: 1% (5 of 537 queries)Open list search: 10% (53 of 537 queries)List search with its own picks: 12% (63 of 537 queries)Brand check: 36% (194 of 537 queries)Fact lookup: 26% (138 of 537 queries)Comparison: 2% (12 of 537 queries)Reviews and forums: 3% (18 of 537 queries)Other: news, official rules, not labeled: 10% (54 of 537 queries) 36% Brand check 26% Fact lookup 12% List withbrands 537 queries
Free
Topic search: 7% (16 of 225 queries)Open list search: 55% (124 of 225 queries)List search with its own picks: 10% (23 of 225 queries)Brand check: 10% (23 of 225 queries)Fact lookup: 8% (17 of 225 queries)Comparison: 2% (5 of 225 queries)Reviews and forums: 1% (2 of 225 queries)Other: news, official rules, not labeled: 7% (15 of 225 queries) 225 queries
Plus
Topic search: 1% (5 of 537 queries)Open list search: 10% (53 of 537 queries)List search with its own picks: 12% (63 of 537 queries)Brand check: 36% (194 of 537 queries)Fact lookup: 26% (138 of 537 queries)Comparison: 2% (12 of 537 queries)Reviews and forums: 3% (18 of 537 queries)Other: news, official rules, not labeled: 10% (54 of 537 queries) 537 queries
Kind Free Plus
Topic search 7% 16 5% without the self-report 1% 5 1% without the self-report
Open list search 55% 124 55% without the self-report 10% 53 14% without the self-report
List search with its own picks 10% 23 16% without the self-report 12% 63 15% without the self-report
Brand check 10% 23 2% without the self-report 36% 194 11% without the self-report
Fact lookup 8% 17 8% without the self-report 26% 138 34% without the self-report
Comparison 2% 5 3% without the self-report 2% 12 4% without the self-report
Reviews and forums 1% 2 2% without the self-report 3% 18 6% without the self-report
Other: news, official rules, not labeled 7% 15 10% without the self-report 10% 54 16% without the self-report

Only the first search step of each answer, before it had seen any results.

Free
Topic search: 5% (6 of 132 queries)Open list search: 55% (73 of 132 queries)List search with its own picks: 16% (21 of 132 queries)Brand check: 2% (2 of 132 queries)Fact lookup: 8% (11 of 132 queries)Comparison: 3% (4 of 132 queries)Reviews and forums: 2% (2 of 132 queries)Other: news, official rules, not labeled: 10% (13 of 132 queries) 55% Open listsearch 16% List withbrands 10% Other 132 queries Topic search: 5% (6 of 132 queries)Open list search: 55% (73 of 132 queries)List search with its own picks: 16% (21 of 132 queries)Brand check: 2% (2 of 132 queries)Fact lookup: 8% (11 of 132 queries)Comparison: 3% (4 of 132 queries)Reviews and forums: 2% (2 of 132 queries)Other: news, official rules, not labeled: 10% (13 of 132 queries) 55% Open listsearch 16% List withbrands 10% Other 132 queries
Plus
Topic search: 1% (2 of 305 queries)Open list search: 14% (42 of 305 queries)List search with its own picks: 15% (47 of 305 queries)Brand check: 11% (33 of 305 queries)Fact lookup: 34% (104 of 305 queries)Comparison: 4% (12 of 305 queries)Reviews and forums: 6% (17 of 305 queries)Other: news, official rules, not labeled: 16% (48 of 305 queries) 34% Factlookup 16% Other 15% List withbrands 305 queries Topic search: 1% (2 of 305 queries)Open list search: 14% (42 of 305 queries)List search with its own picks: 15% (47 of 305 queries)Brand check: 11% (33 of 305 queries)Fact lookup: 34% (104 of 305 queries)Comparison: 4% (12 of 305 queries)Reviews and forums: 6% (17 of 305 queries)Other: news, official rules, not labeled: 16% (48 of 305 queries) 34% Fact lookup 16% Other 15% List withbrands 305 queries
Free
Topic search: 5% (6 of 132 queries)Open list search: 55% (73 of 132 queries)List search with its own picks: 16% (21 of 132 queries)Brand check: 2% (2 of 132 queries)Fact lookup: 8% (11 of 132 queries)Comparison: 3% (4 of 132 queries)Reviews and forums: 2% (2 of 132 queries)Other: news, official rules, not labeled: 10% (13 of 132 queries) 132 queries
Plus
Topic search: 1% (2 of 305 queries)Open list search: 14% (42 of 305 queries)List search with its own picks: 15% (47 of 305 queries)Brand check: 11% (33 of 305 queries)Fact lookup: 34% (104 of 305 queries)Comparison: 4% (12 of 305 queries)Reviews and forums: 6% (17 of 305 queries)Other: news, official rules, not labeled: 16% (48 of 305 queries) 305 queries
Kind Free Plus
Topic search 5% 6 5% without the self-report 1% 2 1% without the self-report
Open list search 55% 73 55% without the self-report 14% 42 14% without the self-report
List search with its own picks 16% 21 16% without the self-report 15% 47 15% without the self-report
Brand check 2% 2 2% without the self-report 11% 33 11% without the self-report
Fact lookup 8% 11 8% without the self-report 34% 104 34% without the self-report
Comparison 3% 4 3% without the self-report 4% 12 4% without the self-report
Reviews and forums 2% 2 2% without the self-report 6% 17 6% without the self-report
Other: news, official rules, not labeled 10% 13 10% without the self-report 16% 48 16% without the self-report

Only the searches after the first, once it had seen results. Answers that searched once have none.

Free
Brand check: 20% (2 of 10 queries)Fact lookup: 60% (6 of 10 queries)Other: news, official rules, not labeled: 20% (2 of 10 queries) 60% Factlookup 20% Brandcheck 20% Other 10 queries Brand check: 20% (2 of 10 queries)Fact lookup: 60% (6 of 10 queries)Other: news, official rules, not labeled: 20% (2 of 10 queries) 60% Fact lookup 20% Brand check 20% Other 10 queries
Plus
List search with its own picks: 4% (5 of 113 queries)Brand check: 60% (68 of 113 queries)Fact lookup: 29% (33 of 113 queries)Reviews and forums: 1% (1 of 113 queries)Other: news, official rules, not labeled: 5% (6 of 113 queries) 60% Brandcheck 29% Factlookup 5% Other 113 queries List search with its own picks: 4% (5 of 113 queries)Brand check: 60% (68 of 113 queries)Fact lookup: 29% (33 of 113 queries)Reviews and forums: 1% (1 of 113 queries)Other: news, official rules, not labeled: 5% (6 of 113 queries) 60% Brand check 29% Fact lookup 5% Other 113 queries
Free
Brand check: 20% (2 of 10 queries)Fact lookup: 60% (6 of 10 queries)Other: news, official rules, not labeled: 20% (2 of 10 queries) 10 queries
Plus
List search with its own picks: 4% (5 of 113 queries)Brand check: 60% (68 of 113 queries)Fact lookup: 29% (33 of 113 queries)Reviews and forums: 1% (1 of 113 queries)Other: news, official rules, not labeled: 5% (6 of 113 queries) 113 queries
Kind Free Plus
List search with its own picks 0% 0 – without the self-report 4% 5 – without the self-report
Brand check 20% 2 – without the self-report 60% 68 – without the self-report
Fact lookup 60% 6 – without the self-report 29% 33 – without the self-report
Reviews and forums 0% 0 – without the self-report 1% 1 – without the self-report
Other: news, official rules, not labeled 20% 2 – without the self-report 5% 6 – without the self-report

What it searches for

Share of answers with at least one query of each type.

“Words” rows are read from the text of the query, so anyone can check them. “Classifier” rows come from the query labels. On the app the numbers include the self-report; the number without it follows where it differs.

Queries per answer
Without the self-report the app count is a floor: hidden steps have no text.
Free 2.3median 2 1.3 · median 1 without the self-report
Plus 5.4median 5 3.1 · median 3 without the self-report
Names a brand or productclassifier
At least one query names a brand or product (publications such as NerdWallet do not count).
“Buffer pricing 2026 Canva pricing Google Ads keyword planner official”
Free 38%38 of 100 30% · 30 of 100 without the self-report
Plus 92%92 of 100 78% · 78 of 100 without the self-report
Asks about price, fees or rateswords
The query contains a word such as price, cost, MSRP, fee, APY, APR, rate or a $ amount.
“site:capitalone.com venture x rewards annual fee travel credit 2026”
Free 37%37 of 100 36% · 36 of 100 without the self-report
Plus 56%56 of 100 53% · 53 of 100 without the self-report
Asks for reviews, Reddit or forumswords
The query contains review, Reddit, forum, Quora, Trustpilot or complaints.
“best women's workout clothes activewear brands leggings sports bras 2026 reviews”
Free 15%15 of 100
Plus 27%27 of 100
Limits a search to one site (site:)words
The query uses the site: operator.
“site:americanexpress.com platinum card annual fee benefits 2026”
Free 0%0 of 100
Plus 17%17 of 100 6% · 6 of 100 without the self-report
Searches inside a brand's own siteclassifier
site: pointing at a domain labeled as a brand's own site.
“site:americanexpress.com platinum card annual fee benefits 2026”
Free 0%0 of 100
Plus 16%16 of 100 5% · 5 of 100 without the self-report

Pages it reads

Opening a page means reading it beyond the title and snippet in the search results.

The last two rows read the overlap both ways: how much of what it cites it had read, and how much of what it read it cites.

The Free stream showed no page reads in this wave. That can mean ChatGPT worked from search results alone, or that the app does not show its reads; the data cannot tell the two apart.

Answers that opened a page
The API's open_page or find_in_page call; a "view" result in the app stream.
Free 0%0 of 100
Plus 15%15 of 100
Pages opened per answer
Free 0.0median 0
Plus 0.5median 0
Cited pages it had opened
The rest were cited from the search results alone: title and snippet.
Free 0%0 of 348 cited
Plus 6%40 of 624 cited
Opened pages it went on to cite
The rest were read but left out of the citations.
Free –
Plus 87%40 of 46 opened

What reaches the answer

What the reader gets: citations, links in the text, the sites behind them and the brands named.

For the same prompt, the two share 23% of their cited sites on average.

Cited pages per answer
Same prompt: Free higher 11, Plus 73, equal 16
Free 3.5median 4
Plus 6.2median 6
Cited sites per answer
Different domains among the cited pages.
Free 2.9median 3
Plus 4.9median 5
Brands and products named
Same prompt: Free higher 40, Plus 29, equal 31
Free 5.5median 5
Plus 5.2median 5
Answer length, characters
Includes the list of searches each prompt asks for.
Free 6,710
Plus 6,474
Links in the answer text
Links ChatGPT writes into the text, apart from its citations.
Answers with a link in the text
Free 76%76 of 100
Plus 45%45 of 100
Linked pages per answer
Same prompt: Free higher 55, Plus 13, equal 32
Free 4.0median 4
Plus 2.1median 0
Linked pages on sites it did not cite
Linked in the text, and no page of that site is among the answer's citations. Mostly the homepages of brands it names.
Free 92%366 of 397 linked
Plus 56%120 of 213 linked
Linked pages that are homepages
A link to the front page of a site rather than to a page about the question.
Free 51%201 of 397 linked
Plus 52%110 of 213 linked
Linked pages never seen in search results
ChatGPT linked a page that did not come back in any of its searches.
Free 92%367 of 397 linked
Plus 61%130 of 213 linked
Free: 334 cited only, 14 cited and linked, 383 linked onlyPlus: 573 cited only, 51 cited and linked, 162 linked only
The kind of sites cited
Share of each surface's cited pages, by the type of site.
Editorial
Free 53%185 of 348
Plus 30%185 of 624
Brand sites
Free 11%40 of 348
Plus 36%226 of 624
Reference and government
Free 16%54 of 348
Plus 21%128 of 624
News
Free 7%26 of 348
Plus 4%24 of 624
Marketplaces
Free 5%18 of 348
Plus 5%34 of 624
Reviews
Free 5%19 of 348
Plus 3%21 of 624
Community (UGC)
Free 2%6 of 348
Plus 0%3 of 624
Other
Free 0%0 of 348
Plus 0%3 of 624
Big platforms: in the results, and cited
How much of what its searches return comes from the big platforms, and how much of that it cites. Each platform's row is its share of the search results, with its results and citations beside it.
Search results from big platforms
Pages from Reddit, Wikipedia, YouTube, Amazon, Quora, Facebook, Instagram, TikTok, LinkedIn, X, Pinterest, Medium, Yelp, Tripadvisor, among all the pages its searches returned.
Free 0%0 of 1505 results
Plus 4%218 of 5137 results
Cited pages from big platforms
The same sites among the pages it cited.
Free 0%0 of 348 cited
Plus 0%3 of 624 cited
Reddit in the results
Free 0%0 results · 0 cited
Plus 3%166 results · 1 cited
Wikipedia in the results
Free 0%0 results · 0 cited
Plus 0%10 results · 0 cited
YouTube in the results
Free 0%0 results · 0 cited
Plus 0%21 results · 0 cited
Amazon in the results
Free 0%0 results · 0 cited
Plus 0%21 results · 2 cited
"Best …" articles: in the results, and cited
How much of what its searches return are list articles, and how much of what it cites. Then who publishes them: each group's share of the list articles in the results, with its results and citations beside it.
Search results that are "best …" articles
List articles among all the pages its searches returned.
Free 62%927 of 1505 results
Plus 25%1262 of 5137 results
Cited pages that are "best …" articles
List articles among the pages it cited.
Free 69%240 of 348 cited
Plus 33%206 of 624 cited
Published by media and review sites
Editorial, review and news sites.
Free 89%823 results · 210 cited
Plus 87%1101 results · 180 cited
Published by brands on their own sites
A brand's own "best …" list, usually with the brand on it.
Free 5%42 results · 8 cited
Plus 7%89 results · 9 cited
Published elsewhere
Marketplaces, community and reference sites, and sites not yet labeled.
Free 7%62 results · 22 cited
Plus 6%72 results · 17 cited
Brands named for the same prompt
727 brands across 100 prompts; on average they agree on 49% per prompt.
Free only: 206Both: 348Plus only: 173

Ads

Ads shown in the answer, where the ad check ran.

Answers with an ad
Only answers where the ad check ran.
Free 45%45 of 100
Plus 0%0 of 100

What the web can move

How much of the answer is open to what search finds, rather than fixed by what the model already knows. Content cannot change the model's memory, but it can change what these searches return.

The rows below are the ways in, each a share that can be checked on the answer pages. Page types and site types come from the classifier's labels. Cited pages that are among the search results the app shows: Free 335 of 348, Plus 622 of 624.

Answer brands the first search did not name
83% Free 436 of 524 brands
47% Plus 246 of 521 brands
The first search runs before any results, so a brand it did not name came later: the web had a chance to put it there, though some may still come from memory. Answers whose first search is known.
Where the web gets in
Each row is one opening for content, with what a brand would do about it.
Answers with an open list search
A "best …" search that names no brand: the pages it returns supply the candidates. Ranking on them is the way in.
Free 73%73 of 100 64% · 64 of 100 without the self-report
Plus 35%35 of 100 34% · 34 of 100 without the self-report
Cited pages that are "best …" articles
List articles it cites. Being on these lists puts a brand in front of it.
Free 69%240 of 348 cited
Plus 33%206 of 624 cited
Answer brands that no search named
They came from the pages it read, not from its queries. Answers that searched.
Free 79%406 of 517 brands 84% · 434 of 517 brands without the self-report
Plus 37%187 of 507 brands 48% · 244 of 507 brands without the self-report
Cited pages from brands' own sites
Product, pricing and help pages that answer its fact lookups and brand checks. A brand controls these pages.
Free 11%40 of 348 cited
Plus 36%226 of 624 cited
Searches that added no cited page
No page it cited first came back in that search step: it looked and used nothing new. A page that answers such a search well meets little competition.
Free 12%12 of 98 steps
Plus 5%6 of 124 steps

Every prompt in detail

Each prompt with its searches, citations, links and brands on all three surfaces is on its category page.

Read with care

What these numbers can and cannot say.

  • The prompts decide most of what these numbers can say. They are 100 commercial searches such as “best credit cards for travel” or “where to open a roth ira”, typed word for word as US searchers type them, in five categories. There are no comparison, advice or opinion prompts. Read every share as true for prompts like these, not for ChatGPT in general.
  • One run per prompt and surface. ChatGPT does not answer the same way twice, so a single prompt can differ by chance; averages over many prompts are what to read.
  • The surfaces run different models (Free: gpt-6; Plus: gpt-6-thinking). A difference between them is the surface and the model together.
  • The app hides the query text of some search steps (Free 5%, Plus 19%). Query numbers for the app include ChatGPT's own list of its searches to fill that gap; a list the model wrote can miss or add a search.
  • Search results are not compared: the API returns every result it got, the app only what it shows.