For AI agents: a documentation index is available at the root level at /llms.txt and /llms-full.txt. Append /llms.txt to any URL for a page-level index, or .md for the markdown version of any page.
April 2026
Multimodal Search & Agent-Friendly Docs
Multimodal search : Search across video (≤120s clips), audio (≤80s clips), and images alongside text in a single query
Gemini Flash cross-modal reranker alongside Voyage text reranker
Multi-query paraphrase retrieval with cross-pipeline deduplication
ARN-based indexing for S3
Agent-friendly documentation : llms.txt, structured API reference for LLMs
YouTube service improvements
March 2026
Odyssey Dataset & OCR
Odyssey private market intelligence launched — companies, investors, people, deals, funds, LPs, service providers, patents, credit analysis
Author field for article dataset search
URIs returned in non-inference query results
OCR functionality for image-based documents
YouTube video indexing
Bounding box data in query results
Cancel and rollback endpoints for indexing jobs
February 2026
Multi-Cloud Indexing & Environment Migration
Azure Blob Storage indexing — full container, single file, and directory
Cloudflare R2 indexing support
URL indexing endpoint
Reranker support for queries
Page numbers returned in search results
Document deletion fixes for grouped file sets
Environment migration endpoint — promote collections from dev → prod without reindexing
Job failure count fixes and atomic operations
January 2026
Billing information in indexing job responses (cost, credits used, processing type)
Two new directory-scoped indexing endpoints for S3 and GCS
Massive RDS connection pooling improvements
Structured SSE streaming with id, event, and data fields
December 2025
V2 API Launch
V2 API overhaul — new RESTful endpoint structure (/v2/collections/...)
Namespace management for collections
Metadata filtering on queries
Custom top_k and custom prompting support
Skip existing files during re-indexing
Supported file types expansion
Groq LLM integration
November 2025
SSE streaming improvements for responses
PostHog analytics integration
Tool calling support — OpenAI-compatible function calling with client-side execution
Response formatting improvements for AI output readability
Docs site launched with Fern
October 2025
Infinite Responses & SDK Compatibility
OpenAI SDK-compatible chat completions endpoint (/v1/chat/completions)
Vercel AI SDK support added
Telemetry support for chat completions
System prompt management for large context processing