Jina AI provides AI-powered APIs for search, embeddings, reranking, web reading, text segmentation, classification, and deep research. Build search, RAG, and AI agent applications with a unified API.
Convert a web page URL into clean, LLM-friendly markdown text using Jina Reader. Supports JavaScript rendering and content extraction.
12 parameters
URL
Return Format
CSS Target Selector
Wait For Selector
Remove Selector
Use Browser Engine
Include Links
Include Images
Generate Alt Text
Skip Cache
Token Budget
Timeout
Search Web
Search the web using Jina Search and get results converted to LLM-friendly text. Each request costs a minimum of 10,000 tokens.
5 parameters
Search Query
Limit to Site
Include Links
Include Images
Skip Cache
Create Embeddings
Convert text into vector embeddings for semantic search, RAG, classification, and clustering using Jina Embeddings.
6 parameters
Input Text
Model
Task
Dimensions
Normalize
Late Chunking
Rerank Documents
Rerank a list of documents by relevance to a query. Used to improve search result ordering for RAG and search applications.
5 parameters
Query
Documents
Model
Top N
Return Documents
Segment Text
Tokenize and segment text into chunks. Useful for understanding token counts and splitting long documents for processing.
7 parameters
Content
Return Chunks
Return Tokens
Max Chunk Length
Tokenizer
Head
Tail
Classify Text
Classify text into categories using zero-shot classification. Provide labels and the API will determine the best match.
3 parameters
Input Text
Labels
Model
Deep Search
Perform deep research on a topic using Jina DeepSearch. An AI agent reasons, searches the web, reads pages, and synthesizes a comprehensive answer. Average response time is about 57 seconds.
6 parameters
Query
Reasoning Effort
Token Budget
Boost Domains
Exclude Domains
Only These Domains
Make HTTP Request
Make an HTTP request to any URL with full control over method, headers, and body.