Sign in
← AI PulseAug 20, 2026

Deep · news · Multi-source brief

xAI Introduces Grok Imagine for Image Generation and Collections for RAG

xAI has launched Grok Imagine for generating images from text prompts and Collections for managing document sets for retrieval-augmented generation (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 Generation | SpaceXAI Docs
Save

xAI has introduced new capabilities for developers, including Grok Imagine for image generation and Collections for document management. Grok Imagine allows users to create images from text prompts, supporting batch generation and control over aspect ratio, resolution, and quality through an API.

Collections provides tools for integrating enterprise requirements and internal knowledge bases with the xAI API. This service is designed for applications such as RAG and searching large document sets, offering persistent document storage with semantic search capabilities.

Key Points

  • Grok Imagine enables image generation from text prompts using the Grok Imagine model.
  • The Grok Imagine API supports batch generation of multiple images.
  • Users can control aspect ratio, resolution, and quality during image generation.
  • Collections offers persistent document storage with semantic search across multiple documents.
  • Collections allows users to upload and search documents by adding them to collections.
  • Metadata fields can be defined for collections, with options like required, unique, and inject_into_chunk.
  • xAI states that user data stored on Collections is not used for model training purposes.

Context

According to xAI documentation, Collections are distinct from direct file attachments in chat messages, providing persistent storage and semantic search across many documents. Users can create two entities within the Collections service: a File, which is a single uploaded file, and a Collection, which groups files with an embedding index for efficient retrieval. When a collection is created, embeddings can be automatically generated for uploaded files, enabling semantic search across multiple collections. A single file can belong to multiple collections.

xAI supports various UTF-8 encoded text files and provides special file conversion and chunking techniques for certain MIME types. The platform also offers a headless mode for scripting and automated tasks, accessible via the grok command-line interface, which can output in formats like JSON or streaming JSON. This mode is suitable for scripts, bots, or other machine-friendly operations.

Why It Matters

The introduction of Grok Imagine and Collections provides developers with expanded capabilities for integrating generative AI and knowledge management into their applications. Builders can leverage Grok Imagine for creative content generation and Collections for enhancing information retrieval and contextual understanding in AI systems, particularly for RAG applications.

What To Do

  • Review the xAI documentation for Grok Imagine to understand the API parameters for image generation.
  • Explore the Collections documentation to learn about creating and managing document collections for RAG.
  • Compare the features of Collections with the Files API to determine the appropriate solution for persistent document storage versus conversational context.
  • Note the supported MIME types for Collections to ensure compatibility with existing document formats.

Keep Exploring

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