Does OAI-SearchBot train OpenAI's models?

No. OpenAI states OAI-SearchBot is used to build the search index behind ChatGPT Search, not for model training. Training collection runs separately through GPTBot, so the two permissions are independent.

I allow OAI-SearchBot but still never appear in ChatGPT Search. Why?

Index inclusion is necessary, not sufficient. ChatGPT Search also leans on Bing's index, so weak Bing coverage hurts; and selection favors pages whose passages directly answer the query. Check Bing Webmaster Tools coverage next.

What Is OAI-SearchBot?

Last updated: July 5, 2026 · Menra Team

OAI-SearchBot is OpenAI's crawler for building and refreshing the search index that powers ChatGPT Search. It appeared in July 2024 alongside the SearchGPT prototype announcement, identifies itself with a user-agent containing OAI-SearchBot plus a documentation URL, and is explicitly separated from OpenAI's training crawler.

Why this bot is the gate to ChatGPT citations

When ChatGPT Search answers a query, its candidate sources come from OpenAI's own index — populated by OAI-SearchBot — layered on Bing's index and publisher partnerships. A site that disallows OAI-SearchBot excludes itself from the OpenAI-indexed portion of that pool, narrowing its path to citations to whatever Bing supplies. OpenAI's own bot documentation frames the token plainly: allow it to appear in search results, block it to opt out. For any brand where ChatGPT answers commercial queries in its category, this single robots.txt line is among the highest-leverage technical settings that exist.

The 2023 blocklist hangover

OAI-SearchBot launched a year after GPTBot, and that sequencing created a real-world problem: thousands of sites had already deployed "block everything OpenAI" robots rules in 2023, then silently inherited invisibility in ChatGPT Search when the product shipped in October 2024. Auditing robots.txt for stale AI-bot rules is now a standard first step in a ChatGPT visibility program — the fix costs one line and pays back within a recrawl cycle.

Technical profile

Respects robots.txt under the OAI-SearchBot token; OpenAI publishes IP ranges for verification.
Crawls without rendering JavaScript, so critical content must exist in server-delivered HTML.
Recrawl cadence scales with site activity; fresh, frequently updated pages get revisited faster.

Recommended configuration

User-agent: OAI-SearchBot
Allow: /

Pair the allow rule with clean XML sitemaps and strong Bing indexing, since the two indexes jointly determine your retrievability. Then verify outcomes: log crawler visits, and track whether ChatGPT actually cites you across your prompt set — the measurement loop described throughout this glossary.

Frequently asked questions

Does OAI-SearchBot train OpenAI's models?: No. OpenAI states OAI-SearchBot is used to build the search index behind ChatGPT Search, not for model training. Training collection runs separately through GPTBot, so the two permissions are independent.
I allow OAI-SearchBot but still never appear in ChatGPT Search. Why?: Index inclusion is necessary, not sufficient. ChatGPT Search also leans on Bing's index, so weak Bing coverage hurts; and selection favors pages whose passages directly answer the query. Check Bing Webmaster Tools coverage next.

Keep exploring

See how AI engines talk about your brand — track mentions across ChatGPT, Perplexity, Claude, Gemini and 5 more. Start with Menra