Agentic Web Search

Perplexity vs Google Custom Search JSON API

PerplexityGoogle Custom Search JSON API

28%

72%

Leading: Google Custom Search JSON API (72.0%)

Insufficient data

This matchup has 25 decisive cases (minimum 30 required for publication).

Metric	Value
Perplexity wins	7
Google Custom Search JSON API wins	18
Abstains (no tool)	1
Other tool chosen	466
Decisive cases	25
Perplexity win rate (unweighted)	28.0%
95% CI	14.3% - 47.6%
Perplexity win rate (weighted)	28.0%

Verified critics can leave comments here.

Verified critics can leave comments here.

Model range	Tier	Perplexity	Google Custom Search JSON API	None	Other	A rate
DeepSeek Flash 3.2–4.1	Mid	0	0	0	27	n/a
DeepSeek Pro 4–4 0813	Frontier	0	0	0	27	n/a
DeepSeek R1	Frontier	0	0	1	26	n/a
Devstral	Mid	0	0	0	18	n/a
Gemini Flash 2.5–3.8	Small	0	0	0	24	n/a
Gemini Pro 2.5–3.1 Preview	Frontier	0	0	0	6	n/a
GLM 5 Turbo–5.3	Frontier	0	0	0	27	n/a
GPT 5.4–6 Astra	Frontier	0	0	0	27	n/a
GPT Coding 5.3 Codex–6.1 Sol	Frontier	0	0	0	27	n/a
GPT Mini 5.4 Mini–6 Luna	Mid	0	0	0	27	n/a
Haiku	Small	6	0	0	21	100%
Kimi 2.5–3	Frontier	0	0	0	27	n/a
Llama Maverick	Frontier	0	9	0	16	0%
Llama Scout	Small	0	9	0	15	0%
MiMo Pro 2–2.6	Frontier	0	0	0	26	n/a
MiniMax M 2.7–3	Frontier	0	0	0	27	n/a
Mistral Small	Mid	0	0	0	27	n/a
Opus 4.6–5.5	Frontier	0	0	0	27	n/a
Qwen Coder	Mid	1	0	0	17	100%
Sonnet 4.6–5.5	Frontier	0	0	0	27	n/a

Prompt	Tier	Perplexity	Google Custom Search JSON API	None	Other	A rate
ai-agent-application	Beginner	5	9	0	152	36%
ai-agent-application	Intermediate	2	8	1	155	20%
ai-agent-application	Advanced	0	1	0	159	0%