Sign in
← AI PulseAug 20, 2026

Wire · news · Multi-source brief

xAI Details Image Understanding for Grok 4.6, Collections for RAG

xAI has provided documentation for image understanding capabilities in its models, including Grok 4.6, allowing images as input for contextual responses, and introduced Collections for managing and querying large document sets for RAG applications.

By Illumora Editorial

Source · Aug 20, 2026, 3:13 PM · On Illumora · Aug 20, 2026, 4:09 PM

Media from the primary source — shown here so you can stay on Illumora.

Synthesized from multiple allowlisted primaries on the same event. Lanes →

Brief drafted by Illumora’s editorial model from the linked primary source. Ops desk reviews flagged pieces. How we write →

Read the source →xAI Docs — Image Understanding | SpaceXAI Docs
Save

xAI has released documentation detailing how its models, such as Grok 4.6, can interpret visual context from images provided as input. This capability allows users to send images alongside text prompts, enabling the model to consider visual information when generating responses. The company also introduced Collections, a service designed to integrate enterprise requirements and internal knowledge bases with the xAI API, particularly for Retrieval Augmented Generation (RAG) applications.

Key Points

  • xAI models, including Grok 4.6, support image input for contextual responses.
  • The request message for image understanding is similar to text-only prompts but uses a list of objects for content.
  • Image URLs can be either base64 encoded strings or public internet URLs.
  • xAI advises against storing request/response history on the server when sending images to prevent potential request failures.
  • Collections provide persistent document storage with semantic search across multiple documents.
  • Collections support various document formats and allow for configuring chunking and embeddings.
  • xAI states that user data stored on Collections is not used for model training purposes.

Context

According to xAI, the image understanding feature allows models to consider image context when generating responses. The primary difference from a text-only prompt is the structure of the content field, which becomes a list of objects containing both image and text input. For managing and querying documents, xAI's Collections service enables users to upload and search through documents by adding them to collections. This service is designed for integrating enterprise data and internal knowledge bases with the xAI API, supporting use cases like RAG applications, as detailed by xAI Docs.

Why It Matters

These capabilities offer developers new ways to build applications that combine visual and textual understanding, and to integrate proprietary data for more informed model responses. The image understanding feature expands the types of input models can process, while Collections provide a structured method for managing and leveraging large datasets for retrieval-augmented generation.

What To Do

  • Review the xAI documentation on image understanding to understand the input message structure.
  • Test sending images as base64 data URLs or public URLs to Grok 4.6.
  • Note the advice to disable storing request/response history on the server when working with images.
  • Explore the Collections API or Console to create and manage document collections for RAG applications.
  • Compare the features of Collections with the Files API for persistent document storage versus direct chat message attachments.

Keep Exploring

/atlas/**grok**-family /techniques/multimodal-grounding /techniques/output-schema /studio?pack=foundation