JinaAI

JinaAI is a search foundation for building applications that need high-quality retrieval and LLM-ready content. It includes core building blocks like multimodal and multilingual embeddings, a reranker to maximize search relevance, and a reader that converts web content into a format easier for language models to consume.

Use r.jina.ai to read a URL and fetch its content in an LLM-friendly way by prefixing the target link (for example, r.jina.ai/http://example.com). This reader-style approach is designed to turn page text into structured input your downstream model can process. For web search workflows, use s.jina.ai to fetch SERP results directly, enabling search within JinaAI-powered flows.

JinaAI also supports a range of request parameters to control how content is fetched and transformed. You can choose a browser engine for the fetch quality/speed tradeoff, adjust response format and timeouts, and manage token budgets to avoid request failures. There are options for returning JSON (including URL, title, content, and timestamp when available), pagination for multi-page documents, and content controls such as excluding elements, stripping images, or converting images with jina-ocr-v1 (noting it costs more tokens).

You can start without a credit card or registration. To raise rate limits, add your API key via API Key & Billing; consult the site’s rate limit table for current values. For privacy and deployment considerations, JinaAI offers TLS encryption in transit and GDPR compliance on the cloud side, plus an open-source framework option for on-premises control. There’s also an experimental EU residency option where infrastructure and processing for a request are located within the EU.

How do I get a JinaAI API key and use it for a higher rate limit?

Create an account, log in, and go to API Key & Billing in your account settings to generate or manage your API key. Include your API key in requests to access a higher rate limit and check the rate limit table for the latest values.

Do I need a credit card or registration to start using JinaAI?

No. You can start instantly without a credit card or registration.

What is r.jina.ai and how do I read a URL with it?

r.jina.ai is a reader-style approach that fetches content from a URL and prepares it as LLM-friendly input. Prefix the URL with r.jina.ai—for example, r.jina.ai/http://example.com.

What is s.jina.ai and when should I use it?

s.jina.ai enables web search by fetching SERP results for use inside your JinaAI-powered workflows. Use it when you need search results rather than a single page’s contents.

Can I control how much content is returned and how requests behave?

Yes. You can set timeouts and token budgets, choose browser engine behavior, and control the returned output via options like JSON response formatting. You can also refine extraction with CSS selectors (extract only, wait for, exclude/remove) and image handling.

Is there an EU residency option for data processing?

Yes—there’s an experimental EU Compliance setting that locates infrastructure and data processing for the request within the EU.