API Returning empty search_results object

[Mike]

Describe the Bug

For a particular domain the API has suddenly stopped returning a search_results object – i.e. there are no citations returned. See:

curl --request POST --url https://api.perplexity.ai/chat/completions --header “accept: application/json” --header “authorization: Bearer pplx-xxx” --header “content-type: application/json” --data “{\“model\”:\“sonar-pro\”,\“messages\”:[{\“role\”:\“user\”,\“content\”:\“What is a TIA \”}],\“search_results\”:true,\“search_domain_filter\”:[\“stroke.org.uk\”]}”

The problem seems to be limited to only this site which has worked fine until now. Nothing else has changed at our end.

Expected Behavior

What you expected to happen.

Actual Behavior

What actually happened.

Steps to Reproduce

1. Call the API with the following request:
2. Observe the unexpected behavior.

API Request & Response (if applicable)

Environment

• API Version: [e.g., sonar-3.1]
• SDK (if applicable): [e.g., Python SDK v0.5]
• Operating System: [e.g., MacOS, Linux, Windows]

Logs or Screenshots (if applicable)

Add any logs or screenshots that can help debug the issue.

Additional Context

Add any other context about the problem here.

[Vasanth_S]

I am also facing an issue with the API. I am getting the Citations field as empty. It was working on 1st August. Any update?

“citations”: ,
“search_results”: ,
“object”: “chat.completion”,

[Mike]

Vasanth - I’ve had email from Perplexity suggesting this might be a robots.txt issue. It isn’t. Not sure what it is. Perhaps it is connected with Cloudflare’s move yesterday to block Perplexity. See Perplexity is using stealth, undeclared crawlers to evade website no-crawl directives

[Igor_Otaviano]

Same problem here, empty search results and citations… It’s frustrating. Perplexity searches work really well on the web, but api is not working as it should.

[vikvang-PPLX-team]

Thanks for bringing this to our attention! Our engineering team is currently investigating this issue. Appreciate your patience!

[jamesliounis]

We will be sharing some additional instructions shortly.

[Mike]

Confirmed Cloudflare issue.

[Mike]

Come on guys, we need some answers please! Even if it’s just simple instructions how to allow Perplexity bot through Cloudflare.

[vikvang-PPLX-team]

Hi! Will get back to you shortly! The team is still working on this. Thank you for your patience!

[Mike]

AWS WAF may have stopped allowing Perplexity through for one of our websites. Has anyone else seen this? Maybe, but not necessarily, related to Cloudflare’s 4 August block on Perplexity.

Adding WAF rule (http.user_agent contains “PerplexityBot”) or (http.user_agent contains “Perplexity-User”) to AWS WAF allowlist has not helped.