Grok Imagine Image 2.0 Guide: Prompts, Editing and Settings

All guides · Grok Imagine Image in Fuser

Quick answer: the Grok Imagine prompts that work best in Grok Imagine Image 2.0 in Fuser are short briefs in plain sentences: name the medium first, describe the scene and light, put any words you want rendered in quotation marks and say where they go, and pick quality: medium for final images. To edit, connect one or more images: the same prompt field becomes an instruction, and with several inputs you can name each one as <IMAGE_0>, <IMAGE_1> and so on, the convention in xAI's API reference. In Fuser this is a single node, Grok Imagine Image, which switches from text-to-image to edit as soon as an image is connected. Eight tested prompts you can copy are in the prompt library below.

What the model is and how Fuser runs it

xAI's current image model is grok-imagine-image-2.0, which it documents for generation and for editing with up to five source images (xAI Imagine overview). Fuser runs both modes, text-to-image and edit, from one node, and you never choose between them. With the Images input empty the node generates from text; with one or more images connected it calls the edit endpoint (Fuser docs).

The name is shared with xAI's video models, so check which node you are on. This guide covers stills only. The video nodes have their own pages: Grok Imagine video and Grok Imagine video edit.

What changed from 1.x. xAI released Grok Imagine Image 2.0 in August 2026, first as Quality Mode in the Grok web and mobile apps (TestingCatalog), then in its API as grok-imagine-image-2.0. The gains you will notice are small, dense text that stays sharp, layouts planned around the type, and edits that combine up to five reference images. The quality setting is new as well: xAI documents it as supported only on 2.0 (xAI image generation docs). On Artificial Analysis' text-to-image arena, 2.0 was 6th at 1,166 Elo when we checked on 23 October 2026, against 1,049 for the earlier Grok Imagine quality model and 1,022 for the original. The Grok Imagine Image node in Fuser runs 2.0, so every prompt in this guide uses it.

How to write a Grok Imagine Image prompt

A prompt can be long, so there is no need to compress it into a list of tags. What worked in our runs was a short brief in this order:

1. Subject and medium. "Studio product photo of a matte sage-green ceramic coffee mug", or "Minimal gig poster for a jazz night".

2. Setting and props. "On a pale oak table, a single sprig of rosemary beside it."

3. Light and lens. "Soft window light from the left, shallow depth of field."

4. Exact text, in quotes, with a position. "The mug has the word "SLOW" printed in small black serif capitals."

5. Style or layout. "Swiss poster layout, subtle paper grain."

xAI lists the styles the model covers as "ultra-realistic photography to anime, oil paintings, and pencil sketches" (xAI image editing docs), so name the medium in the first few words rather than leaving it to chance.

Grok Imagine prompts you can copy

xAI's Imagine app organises its templates around jobs: photo edits, product colour changes, e-commerce shots, professional headshots, icons, character sprites, emojis and merchandise (TestingCatalog). We wrote one prompt for most of those jobs and ran each once in Fuser's Grok Imagine Image node at 1K, quality medium and 3:4. Paste a prompt as it is, then swap the subject, the words in quotes and the accent colour.

1. Poster with exact type. "Exhibition poster on bone-white uncoated paper, strict Swiss grid. The top two thirds is a black-and-white photograph of a single tall stoneware vase with a cracked ash glaze, standing on a plinth against a plaster wall, lit by soft daylight from the left. The bottom third is a solid terracotta block. Inside it, flush left, the headline "FORM & FIRE" in tight condensed bone-white sans-serif capitals. Below it in small bone-white caps: "CERAMICS 1960–2026 · 12 MAR – 30 JUN · GALLERY TWO". Thin grid rules, subtle paper grain, nothing else on the page." Every word, dash and middot came out right. The thin grid rules also ran across the photograph, so drop that phrase if you want a clean image area.

2. E-commerce product shot. "E-commerce product photograph on seamless pale-grey paper: an unbranded compact transistor radio in butter-yellow plastic with a perforated brushed-aluminium speaker grille, one round cream tuning dial and a short leather carry strap. Three-quarter view, centred, soft north-window light from the left, a gentle soft shadow falling to the right. Medium-format film look, calm and plain. No logos, no text." A clean catalogue frame with no stray branding.

3. Product colour change (edit). Wire the product shot into the node's Images input, then: "Change the radio's plastic body and tuning dial to a deep cobalt blue. Keep the shape, the brushed-aluminium grille, the leather strap, the pale-grey background, the lighting and the shadow exactly as they are." Only the body and dial changed; the grille, strap and shadow stayed in place.

4. Editorial portrait instead of a stock headshot. "Editorial portrait of a fictional ceramicist in her sixties: short silver hair, a faded indigo linen work shirt with a dusting of dry clay on the sleeves, looking straight into the lens with a calm half-smile. Plain warm plaster wall behind her. Soft daylight from a large window at camera left. Head and shoulders, medium-format film, shallow depth of field, natural skin texture, no retouching." Skin texture stayed natural and the clay dust landed on the sleeves as asked.

5. Icon set. "A set of six app icons arranged in a neat 2-column by 3-row grid on a flat cream background: a kettle, a pocket calculator, a desk lamp, a pair of scissors, a teacup and a stopwatch. Each icon is a rounded square tile with the object modelled in soft matte clay-like 3D, in cream and pale grey with a single terracotta accent detail per icon. Identical top-left lighting and soft drop shadows across all six, consistent scale and corner radius. No text." All six tiles share the same light, scale and corner radius.

6. Game asset: character sprite sheet. "Character sprite sheet for a 2D platform game: a small round fox courier wearing a mustard-yellow raincoat and a brown satchel, in clean 32-bit pixel art. Top row: four frames of a walk cycle facing right. Middle row: idle, jump and wave poses. Bottom row: the satchel, a letter and a raincoat hood as separate item sprites. Flat pale-blue background, one consistent limited palette, crisp pixel edges, even spacing, no text." The character stays consistent across every pose, but the four walk frames are almost identical, so use it as a character sheet and draw the in-between frames separately.

7. Merchandise mockup. "Merchandise mockup: flat-lay photograph from directly above of a natural unbleached canvas tote bag on a pale oak table. The bag is screen-printed in two colours, cobalt blue and lemon yellow, with a bold illustration of a halved lemon and, beneath it, the words "CITRUS MARKET" in hand-lettered cobalt capitals and a smaller line "SATURDAYS 7–1". Soft overhead daylight, the canvas weave visible through the ink, the handles laid neatly to the sides." Both lines are spelled correctly; the hand-lettered style drew the I in CITRUS as a small-cap i.

8. Multi-reference scene (edit). Connect the portrait, the cobalt radio and the tote, in that order, then: "The woman from <IMAGE_0> stands in her sunlit pottery studio, smiling, with the tote bag from <IMAGE_2> hanging from her shoulder, its print facing the camera. Behind her, the radio from <IMAGE_1> sits on a wooden shelf among unglazed pots. Soft daylight from a tall window on the left, plaster walls, medium-format film look. Keep her face, the radio and the tote print unchanged." All three references landed where the prompt put them, and the tote's print stayed readable.

Text in images

Text is where this model earns a place in a workflow. Our poster prompt split the layout into regions: "Top third: the headline "LATE NIGHT BLUE" in tall condensed white sans-serif letters... Bottom: "FRI 14 NOV · 9PM · THE CELLAR" in small white caps." Both the 1k and 2k runs rendered every word and the middots correctly and kept the three-band layout.

Two habits made the difference. Quote the exact string, and describe the type (condensed sans, small serif caps) as well as where it sits. Our tests used one headline and one info line per image; we did not test paragraphs of copy. For body copy or anything with many lines, generate the image with space for it and set the type afterwards in Fuser's Compositor.

Settings that matter

The Fuser node exposes four controls besides the prompt and images.

• Aspect ratio. 13 fixed ratios from 2:1 to 1:2, including 20:9, 19.5:9, 9:19.5 and 9:20 for phone screens, plus Auto. In text-to-image, Auto gives 1:1; in edit mode it keeps the ratio of the first input image (Fuser docs).

• Resolution. 1K (default) or 2K. In our runs a 16:9 image at 1k came back at 1280 × 720, a 2:3 image at 1k at 832 × 1248, and the same 2:3 prompt at 2k at 1664 × 2496.

• Quality. Low or medium, default medium in Fuser. Note that xAI's own API defaults to "auto", which it says currently means low for generation and medium for editing (xAI image generation docs), so results from other tools that call the API directly may have been made at low.

• Images. Optional inputs that switch the node into edit mode.

In our side-by-side the gap was small: both settings kept the composition idea and rendered SLOW cleanly, and medium gave slightly smoother surface detail. In Fuser, medium costs more per image than low, 2K costs more than 1K, and edits add a small charge for each input image. Use low for exploring layouts and medium for anything you keep.

Grok image editor: editing a single image

Connect an image and write the change as an instruction. Say what should change and, just as clearly, what must stay: "Move the mug outdoors onto a mossy granite rock beside a misty mountain lake at sunrise. Keep the mug's shape, sage-green colour and the word "SLOW" exactly as they are. Replace the rosemary with a small pine cone." The result kept the mug, its colour and the printed word, and swapped the prop as asked.

Style changes work the same way. "Redraw this image as a loose graphite pencil sketch on off-white paper, visible hatching, no colour except a light sage wash on the mug" kept the composition and the lettering while changing the medium entirely.

For a sequence of changes, xAI recommends chaining: use each output as the input for the next edit (xAI image editing docs). On a Fuser canvas that is a row of Grok Imagine Image nodes, each wired to the one before, so you can go back and change step two without redoing step one.

Combining several images

The edit endpoint accepts up to five images (xAI Imagine overview). xAI's API reference says that when several images are provided you refer to them as <IMAGE_0>, <IMAGE_1> and so on in the prompt. We tested that convention with the mug as the first image and the poster as the second: "A cosy jazz bar at night. The poster from <IMAGE_1> hangs framed on the brick wall. In the foreground on a dark wooden bar counter sits the mug from <IMAGE_0>, unchanged, with steam rising."

Both references landed where the prompt put them, and the poster's text stayed readable in the new scene. Because we set the aspect ratio to 3:2, the output came back at 1248 × 832 rather than matching the first input. In Fuser, the node sends the images on its Images input to the model as one ordered list. If a result swaps two references, swap the tags in the prompt or reconnect the images.

Fix what goes wrong

• Misspelled text: quote the exact string, shorten it, and say where it goes and what type it uses.

• The edit changed things you wanted kept: list them explicitly with "keep ... exactly as they are", as in the mug edit above.

• Wrong image used in a multi-image edit: use the <IMAGE_n> tags rather than "the first photo", and check the input order.

• Edit output has the wrong shape: Auto follows the first input image. Pick a fixed aspect ratio if you need a different frame.

• Drafts look rougher than expected: check the quality setting. xAI's API defaults generation to low; the Fuser node defaults to medium.

Build it as a workflow

The hero image above is the whole pattern on one canvas: two text-to-image nodes make the product and the poster, and a third node combines them. From there the still can go to an upscaler such as Topaz, into the Compositor for final type, or into an image-to-video model like Kling 3.0. To see how Grok Imagine Image compares with other editors, read the best AI image editing models and the best AI models for text in images.