mudpie

original analysis · 3 min read

.com still leads .ai across nine YC cohorts

A current snapshot of listed website TLDs across 1,586 YC directory rows, with .com and .ai counts kept separate from launch-time history or adoption claims.

Published · Updated

The current YC directory’s largest listed suffix is .com. The interesting part is how common .ai is in the current cross-section without turning the domain name into a proxy for what a company actually does.

I counted the current listed websites for 1,586 companies across nine YC directory cohorts: Summer 2024 through Summer 2026, including the newer Spring and Fall labels. I normalized the hostname, removed one leading www, and counted the final hostname label. I did not fetch the websites. I did not infer what a company builds from its domain.

There were 1,584 listed websites with a syntactically valid public-domain hostname. One row had no website; another supplied a single-label hostname with no TLD. A URL parser accepts the latter, but it does not belong in a public-domain count. The eligible rows contain 1,584 unique normalized hosts. This checks syntax, not whether every destination currently loads.

Cohort Companies Valid websites .com .ai .io .dev
Summer 2024 248 248 121 68 9 4
Fall 2024 94 94 42 26 2 1
Winter 2025 165 165 80 52 5 8
Spring 2025 143 143 65 41 6 6
Summer 2025 166 166 72 61 4 3
Fall 2025 146 146 62 46 7 8
Winter 2026 199 197 87 57 10 6
Spring 2026 193 193 96 52 6 7
Summer 2026 232 232 109 61 8 7

Across the eligible set, .com accounts for 734 of 1,584 listed websites, or 46.34%. .ai accounts for 464, or 29.29%. .io is 57. .dev is 50.

In this snapshot, .ai is a strong second choice. It is not a clean march toward .ai, and this table is not a historical adoption series.

Summer 2024 had 121 .com companies and 68 .ai companies. Summer 2025 had 72 .com and 61 .ai. In the current Summer 2026 snapshot, the counts are 109 .com and 61 .ai. The .ai count is flat between the S25 and S26 snapshots while the S26 directory is larger; that is a cross-section comparison, not a claim that .ai did or did not grow historically.

The table is descriptive, not explanatory: it records the current listed URL, not why a founder chose it.

OpenTag is a Summer 2026 B2B AI coworker with the listed website tryopentag.com. Experiential Labs is listed at experientiallabs.ai. Caution is listed at caution.co.

Those three rows show why a suffix count is not a business category. They are current listed URLs, not evidence of why the founders chose them. They do not prove traction, technical quality, or agent readiness.

These rows tell us where companies are currently listed. They do not tell us whether the suffix helped them win.

This is current listed-domain metadata, not a historical adoption study. A current YC row may reflect a rebrand, a redirect, a parked domain, or a website chosen after launch. The dataset deliberately does not claim what domain a company used at acceptance, Demo Day, or founding.

The useful signal is narrower: among the companies YC currently lists, .com remains the plurality and .ai is the strongest alternative. That is enough to describe the current listed-domain landscape. It is not enough to predict who wins.

Dataset: /research/yc-cohorts-2026-09-19.json and /research/yc-domain-stats-2026-09-19.json.

About the author

I cofound Lazyweb and publish Mudpie. This is an owner-written publication, not an independent testing organization. Research notes distinguish observations, sourced reporting and editorial judgment.

First1000 ↗ · X ↗