Grok Imagine Image 2.0: AI Image Generator Online
Get Started Free

Grok Imagine Image 2.0

Next-generation Grok image model for text-to-image and prompt-guided editing at 1K and 2K

Category
Image
Modality
Text → Image
Context
—
Released
Aug 2026
Run this model Estimate cost
Strengths

What it's the best tool for

  • Text-to-image and prompt-based editing of an existing picture in one model
  • Two output resolutions, 1K and 2K, with 13 aspect ratios in each
  • Vertical, square and wide framing up to 20:9, 1:2 and 9:19.5
  • Dedicated output quality parameter, JPG, PNG and WEBP formats
  • Up to 20 images per request for fast batch exploration
  • Runs online in the browser with no installation required
Limitations

When to reach for something else

  • Only one reference image per request, multi-reference input is not supported
  • Negative prompts are not supported, unwanted details have to be handled in the main prompt
  • A single output quality mode is available for now
  • Preview status: parameters and model behaviour may still change
Sample output

How Grok Imagine Image 2.0 responds

Prompt
A rain-soaked Tokyo alley at night: neon signage reflected in the puddles, a lone figure under a translucent umbrella, cinematic wide shot. The same frame rendered at 1K and at 2K.
Grok Imagine Image 2.0
Where teams use it

Four scenarios where it pays for itself

01
Banners and ad creatives
One visual delivered in square, wide and vertical framing for every placement
02
Prompt-based photo editing
Change background, lighting and details on an existing shot without an editor
03
Covers and posters
2K frames for websites, decks and printed materials
04
Concepts and mockups
Up to 20 options per request to pick a direction fast
About model

More about Grok Imagine Image 2.0

Grok Imagine Image 2.0: AI Image Generator Online

Grok Imagine Image 2.0 is the new generation image model from xAI, released on 10 August 2026. It handles two jobs inside a single model: building a picture from a text description and editing an existing image through a prompt. The first concept and the follow-up fix happen in the same place, so nothing has to travel between separate services.

Resolutions and framing

Two output resolutions are available, 1K and 2K, each with thirteen aspect ratios — from the familiar 1:1 and 16:9 to elongated 20:9, 1:2 and 9:19.5. That range covers practically every placement: a square for a feed, a wide frame for a website hero, a tall crop for stories and mobile banners. Output quality is controlled by a dedicated parameter, and the file comes back as JPG, PNG or WEBP, ready to drop straight into a layout. A single request returns up to twenty images, so exploration happens in batches instead of one attempt at a time.

Editing by prompt

The image-to-image mode takes a source picture plus a written instruction. That is how you change a background, relight a scene, swap clothing or add and remove objects without opening an editor and masking by hand. One reference image goes in per request, which makes the mode a precise tool for refining a specific frame rather than blending many sources at once.

Getting started

Grok Imagine Image 2.0 runs on NetRoom straight in the browser, with no installation and no keys to manage. Write the prompt in plain language: the model reads scene description, style, lighting and composition. The per-image rate is shown in a separate block on this page.

Who it fits

Designers and social media teams get fast visuals for posts and banners. Marketers get creatives in several aspect ratios at once for testing. Product teams get concepts and mockups before anything reaches production design. The model carries Preview status, so its parameters and behaviour may still change as xAI keeps refining it.

Recent changes

What changed Grok Imagine Image 2.0

  • + Added the Grok Imagine Image 2.0 (xAI) image model to the generation catalog.
Full changelog →

Try Grok Imagine Image 2.0
right now

Free access to basic models. No card, no obligations.