Technical documentation is getting a new primary reader: AI. On Mintlify’s platform, nearly 69% of September traffic is now coming from agents — turning documentation from something humans read into infrastructure machines use to do work.
WHAT’S HAPPENING
Mintlify, a platform used to host developer documentation, says AI agents now account for roughly 68.9% of September traffic across its network.
That represents about 262 million agent requests compared with 118 million human visits.
Mintlify also reports that 70% of the documentation sites on its platform now receive more machine requests than human page loads.
Earlier this year, agent and human traffic were nearly even.
The balance has shifted quickly.
WHY IT MATTERS
Documentation used to be written primarily for a person opening a browser and reading instructions.
AI coding agents work differently.
They search documentation, retrieve API instructions, examine technical references and use that information while writing code or completing tasks.
That means documentation is becoming part of the infrastructure AI agents depend on to operate.
WHO BENEFITS
Software companies with accurate, structured and machine-readable documentation could make it easier for AI agents to understand and use their products.
Developers may also spend less time manually searching documentation when an agent can retrieve the information directly while working.
WHO LOSES
Companies with outdated, incomplete or poorly structured documentation could become harder for AI agents to work with.
Traditional website analytics can also miss much of this activity because many agents do not load pages or execute tracking scripts the same way human browsers do.
WHAT HAPPENS NEXT
Documentation may increasingly need to serve two audiences at once:
Humans who want clear explanations.
AI agents that need structured, complete information they can retrieve automatically.
The change is already underway.
The manual is no longer being written only for people.
The machines are reading it too.
