Agent directory / GPTBot
GPTBot
Observed live across sites Botfall meters — measured, not copied from docs
Last seen11 hours ago
First seen59 days ago
Fetches · 24h1
Fetches · 30d594
Sites seen on3
Identity verified 19%
1,406 requests claimed this identity from addresses outside OpenAI's published ranges — impersonation, not this agent.
Networks observed EUGameHost-LTD EUGAMEHOST LTD · GOOGLE-CLOUD-PLATFORM - Google LLC · HOSTODO - Hostodo · INFRALY-LLC - Infraly, LLC · MICROSOFT-CORP-MSN-AS-BLOCK - Microsoft Corporation · NextGenWebs-NL NextGenWebs, S.L.
| Operator | OpenAI |
| Purpose | AI · training crawl |
| Identifies as | User-agent substring: gptbot |
What the operator publishes
Their documentation, not our measurements — re-checked automatically every day.
| Verify identity by | Published IP ranges — a user-agent alone proves nothing; anyone can send this string. |
| Published IP ranges | https://openai.com/gptbot.json — 18 prefixes listed link checked 2026-09-30 |
| robots.txt stance | Operator documents that it honors robots.txt |
| Operator documentation | https://platform.openai.com/docs/bots link checked 2026-09-30 |
Provenance
What establishes this record, and when it was last checked — no speculative entries.
- Observed in live storefront telemetry (classifier lineage)checked 2026-07-30
Want these numbers for your own site — which agents read it, what they took, and what it cost? Get started free →