--- language: en title: "OAI-SearchBot" description: "OAI-SearchBot crawls pages for ChatGPT search features. Learn about access controls and why allowing the bot does not guarantee a citation." canonical: "https://tropk.ai/en/resources/glossary/oai-searchbot" last_modified: "2026-10-12T16:11:31.890Z" --- # OAI-SearchBot Reviewed on 2026-10-06. ## What it means OAI-SearchBot is OpenAI’s [crawler](/en/resources/glossary/crawler) for presenting websites in [ChatGPT](/en/resources/glossary/chatgpt) search results. Its control in [robots.txt](/en/resources/glossary/robots-txt) is independent of [GPTBot](/en/resources/glossary/gptbot), which is used for training. According to the documentation, sites that disallow OAI-SearchBot do not appear in ChatGPT search responses, though they may still appear as navigation links. ## Why it matters for marketing For an [SEO](/en/resources/glossary/seo) team, allowing this bot is an access condition to check when relevant pages are not retrieved by ChatGPT search. The investigation should include server protections and content delivery network rules. An allow rule in the file does not resolve a block applied at another layer of the infrastructure. ## Example in practice Suppose a store publishes an updated technical support page and allows OAI-SearchBot in robots.txt. The logs show that requests from this bot receive HTTP 403 due to a site protection rule. The team requests a review of the block and monitors subsequent responses. The identified issue is page access, without attributing the failure to editorial quality. ## How to apply or check Check the specific rule for OAI-SearchBot and compare the source IPs with the official list published by OpenAI before adjusting permissions. Check paths, response codes, and delivered content. The documentation gives an approximate 24-hour timeframe for adapting to a change in robots.txt; record the time and avoid judging the effect immediately. ## Important distinctions A successful visit does not guarantee [indexing](/en/resources/glossary/indexing), ranking, or [citation](/en/resources/glossary/ai-citation) for any query. Navigation links also do not mean that a page was used as a source in the response. OAI-SearchBot handles automated [crawling](/en/resources/glossary/crawling) for search; its rule does not replace the decision about training with GPTBot or describe all the access users may request through ChatGPT. ## Sources and references - [Overview of OpenAI Crawlers]() — OpenAI ## Continue a leitura - [GPTBot]() - [Indexing]() - [robots.txt]() ## About the author [Vinicius Pacheco]() Responsible for Tropk’s technical content, focusing on crawling, indexing, structured data, and agent access to content. --- [All pages in Markdown]()