Gemini vs ChatGPT: where they really get their answers

Where they diverge

The biggest gaps this month: whatcar.com (Gemini 0.0% vs ChatGPT 1.1%); youtube.com (Gemini 1.1% vs ChatGPT 0.1%); mercedes-benz.com (Gemini 0.0% vs ChatGPT 0.9%); gob.pe (Gemini 0.1% vs ChatGPT 0.9%); which.co.uk (Gemini 0.2% vs ChatGPT 0.9%).

Figure 1Share gap: Gemini minus ChatGPT

Data table

Series	gap
whatcar.com	-1.1pp
youtube.com	+1pp
mercedes-benz.com	-0.9pp
gob.pe	-0.8pp
which.co.uk	-0.7pp
kia.com	-0.7pp
fda.gov	-0.7pp
emirates.com	-0.7pp

Where they agree

1 sources sit in both models' top ten: which.co.uk.

Gemini's top sources

youtube.com 1.1%; reddit.com 0.9%; carwow.co.uk 0.4%; wikipedia.org 0.3%; theguardian.com 0.3%.

Figure 2Gemini — top sources

Data table

Series	share
youtube.com	1.1%
reddit.com	0.9%
carwow.co.uk	0.4%
wikipedia.org	0.3%
theguardian.com	0.3%
moneysavingexpert.com	0.3%
nih.gov	0.2%
google.com	0.2%
which.co.uk	0.2%
trustpilot.com	0.2%

ChatGPT's top sources

whatcar.com 1.1%; gob.pe 0.9%; which.co.uk 0.9%; mercedes-benz.com 0.9%; gov.uk 0.8%.

Figure 3ChatGPT — top sources

Data table

Series	share
whatcar.com	1.1%
gob.pe	0.9%
which.co.uk	0.9%
mercedes-benz.com	0.9%
gov.uk	0.8%
kia.com	0.8%
fda.gov	0.8%
hyundai.com	0.8%
gov.sa	0.7%
tesla.com	0.7%

Do they rank brands the same way?

Across recent analyses, Gemini and ChatGPT rank shared brands with a median correlation of 0.40; the median across all benchmarked model pairs is 0.41.

Figure 4Brand-ranking agreement by model pair

Data table

Series	median rho
Google AIO × Google AI Mode	0.6
Perplexity × ChatGPT	0.5
Gemini × Google AI Mode	0.5
Gemini × Google AIO	0.5
ChatGPT × Google AIO	0.4
Claude × Gemini	0.4
Perplexity × Meta AI	0.4
Gemini × DeepSeek	0.4
DeepSeek × Meta AI	0.4
Perplexity × Gemini	0.4
Perplexity × DeepSeek	0.4
Perplexity × Google AI Mode	0.4
Grok × Google AI Mode	0.4
ChatGPT × Meta AI	0.4
DeepSeek × Google AIO	0.4
Claude × Meta AI	0.4
ChatGPT × Google AI Mode	0.4
Perplexity × Google AIO	0.4
Gemini × ChatGPT	0.4
Gemini × Grok	0.4
Gemini × Meta AI	0.4
DeepSeek × Google AI Mode	0.4
ChatGPT × Grok	0.4
ChatGPT × DeepSeek	0.4
Grok × Google AIO	0.4
Claude × Google AI Mode	0.4
Claude × Perplexity	0.4
Google AI Mode × Meta AI	0.4
Claude × ChatGPT	0.4
Perplexity × Grok	0.4
Grok × DeepSeek	0.4
Claude × Google AIO	0.4
Claude × Grok	0.3
Google AIO × Meta AI	0.3
Claude × DeepSeek	0.3

Source mix

Share of citations by source category, Gemini vs ChatGPT.

Figure 5Source mix: the tale of the tape

Data table

Series	social platforms	news publishers	comparison & review sites	academic & research	marketplaces	B2B research	everything else
Gemini	2.4%	1.4%	1.4%	0.4%	1.2%	0.5%	92.6%
ChatGPT	0.9%	2.3%	3.2%	0.4%	1.1%	1.6%	90.5%

Method & limitations

• Measured across 63,430 citations captured by the platform in July 2026 — a large, continuously refreshed sample, not a census of AI answers.
• Only domains cited across at least 2 independent projects are named.
• Source preferences reflect the questions asked as well as each model’s retrieval behaviour — differences are descriptive, not a quality ranking.

How every figure on this page is counted, floored and reviewed: our methodology.

More from Model vs Model

• Claude vs ChatGPT: Where They Get Their Answers — July 2026
• Grok vs Google AI Mode: Where They Get Their Answers — May 2026
• Google AIO vs Google AI Mode: Where They Get Their Answers — June 2026

The full Model vs Model archive