Introduction
You can generate and manage API keys from your Cortex Dashboard. All endpoints require an API key sent as a Bearer token in the Authorization header.
Welcome to the Cortex SDK API Reference. This section documents every available endpoint and how to interact with them to power AI apps and agents with intelligent memory and retrieval.
Base URL:
https://api.usecortex.aiContact us to get your API key at founders@usecortex.ai
Upload APIs
Send knowledge into Cortex. Upload documents, batch files, and scrape webpages.
/upload/upload_document/upload/batch_upload/upload/scrape_webpage
Search API
Ask questions. Retrieve personalized answers from memory and indexed sources.
/search/qna— supports an optionalmetadataparameter to filter sources bysource_titleorsource_type.
Listing APIs
Browse what’s in your Cortex memory or knowledge store.
/list/sources/list/sources_by_id/fetch/fetch_content
Knowledge Management
Manage knowledge entries, delete outdated or unwanted context.
/delete_memory
Processing Verification
Check if a file or page was successfully processed.
/upload/verify_processing
💡 Best Practices
- Use sub-tenants to support multi-user isolation in B2B use cases.
- Leverage metadata (e.g.,
source_title,source_typein the QnA API) for fine-grained filtering and agentic retrieval. - Tune search_alpha and recency_bias to control relevance.
- Enable highlight_chunks for chunk-level citations.
Introduction
You can generate and manage API keys from your Cortex Dashboard. All endpoints require an API key sent as a Bearer token in the Authorization header.
Welcome to the Cortex SDK API Reference. This section documents every available endpoint and how to interact with them to power AI apps and agents with intelligent memory and retrieval.
Base URL:
https://api.usecortex.aiContact us to get your API key at founders@usecortex.ai
Upload APIs
Send knowledge into Cortex. Upload documents, batch files, and scrape webpages.
/upload/upload_document/upload/batch_upload/upload/scrape_webpage
Search API
Ask questions. Retrieve personalized answers from memory and indexed sources.
/search/qna— supports an optionalmetadataparameter to filter sources bysource_titleorsource_type.
Listing APIs
Browse what’s in your Cortex memory or knowledge store.
/list/sources/list/sources_by_id/fetch/fetch_content
Knowledge Management
Manage knowledge entries, delete outdated or unwanted context.
/delete_memory
Processing Verification
Check if a file or page was successfully processed.
/upload/verify_processing
💡 Best Practices
- Use sub-tenants to support multi-user isolation in B2B use cases.
- Leverage metadata (e.g.,
source_title,source_typein the QnA API) for fine-grained filtering and agentic retrieval. - Tune search_alpha and recency_bias to control relevance.
- Enable highlight_chunks for chunk-level citations.