Models
xAI
Grok Imagine Image

xAI

llms.txt

Grok Imagine Image

AI Image Editing Model

Mode:

All Text to Image Image Edit

Image

An image generation model by xAI designed to generate highly aesthetic images from text descriptions.

Grok Imagine Image generated image of A vast, dimly lit ancient library carved into a cliffside at golden hour, wit...

Try Model Benchmark examples

2048 x 2048

Max Resolution

Supported Modes

Text to Image Image Edit

Active

Details

Model ID

grok-imagine-image

Also known as: grok-imagine

Creator

xAI

Family

grok

Released

January 2026

Max Input Images

Ready to integrate?

Access grok-imagine-image via our unified API.

Create Account

Available at 2 providers

Starting from

$0.020 /image via fal.ai, xAI

Prices shown are in USD

See all providers

Provider Performance

Fastest generation through xai at 7,683ms median latency with 62.6% success rate.

Aggregated from real API requests over the last 30 days.

Generation Time

xai

p95: 11,909ms

fal

p95: 25,311ms

Success Rate

xai

62.6%

189 / 302 requests

fal

90.7%

262 / 289 requests

Time to First Byte

xai

7,068ms

p95: 10,824ms

fal

7,793ms

p95: 25,314ms

Provider Rankings

#	Provider	p50 Gen Time	p95 Gen Time	Success Rate	TTFB (p50)
1	xai	7,683ms	11,909ms	62.6%	7,068ms
2	fal	7,792ms	25,311ms	90.7%	7,793ms

Data updated every 15 minutes. Based on all API requests through Lumenfall over the last 30 days.

Providers & Pricing (2)

Grok Imagine Image is available from 2 providers, with per-image pricing starting at $0.02 through fal.ai.

fal.ai

Image Edit

fal/grok-imagine-image-edit

Provider Model ID: xai/grok-imagine-image/edit

$0.0020 /input image

View pricing details

xAI

Text to Image Image Edit

xai/grok-imagine-image

Provider Model ID: grok-imagine-image

$0.0020 /input image

View pricing details

grok-imagine API OpenAI-compatible

Integrate Grok Imagine Image into your workflow via the Lumenfall OpenAI-compatible API to programmatically generate high-fidelity images and perform precise image edits. Use a single endpoint to access xAI's image generation technology using standard text-to-image parameters.

Base URL

https://api.lumenfall.ai/openai/v1

Model

grok-imagine-image

Code Examples

Image Edit

/v1/images/edits

curl -X POST \

  https://api.lumenfall.ai/openai/v1/images/edits \

  -H "Authorization: Bearer $LUMENFALL_API_KEY" \

  -F "model=grok-imagine-image" \

  -F "[email protected]" \

  -F "prompt=Add a starry night sky to this image" \

  -F "size=1024x1024"

# Response:

# { "created": 1234567890, "data": [{ "url": "https://...", "revised_prompt": "..." }] }

import OpenAI from 'openai';

import fs from 'fs';

const client = new OpenAI({

  apiKey: 'YOUR_API_KEY',

  baseURL: 'https://api.lumenfall.ai/openai/v1'

});

const response = await client.images.edit({

  model: 'grok-imagine-image',

  image: fs.createReadStream('source.png'),

  prompt: 'Add a starry night sky to this image',

  size: '1024x1024'

});

// { created: 1234567890, data: [{ url: "https://...", revised_prompt: "..." }] }

console.log(response.data[0].url);

from openai import OpenAI

client = OpenAI(

    api_key="YOUR_API_KEY",

    base_url="https://api.lumenfall.ai/openai/v1"

response = client.images.edit(

    model="grok-imagine-image",

    image=open("source.png", "rb"),

    prompt="Add a starry night sky to this image",

    size="1024x1024"

# { created: 1234567890, data: [{ url: "https://...", revised_prompt: "..." }] }

print(response.data[0].url)

Parameter Reference

Required Supported Not available

Core Parameters

Parameter	Type	Description	Modes
`prompt`	string	Required. Text prompt for image generation	T2I Edit

Size & Layout

Parameter	Type	Description	Modes
`size`	string	Image dimensions as WxH pixels (e.g. "1024x1024") or aspect ratio (e.g. "16:9") WxH determines both shape and scale (aspect_ratio and resolution are ignored when size is provided). W:H format is equivalent to aspect_ratio.	T2I Edit
`aspect_ratio`	string	Aspect ratio of the output image (e.g. "16:9", "1:1") Controls shape independently of scale. Use with resolution to control both. If size is also provided, size takes precedence. Any ratio is accepted and mapped to the nearest supported value.	T2I Edit
`resolution`	string	Output resolution tier (e.g. "1K", "4K") `auto` `1k` `1K` `2k` `2K` Controls scale independently of shape. Higher tiers produce larger images and cost more. If size is also provided, size takes precedence for scale. Any tier is accepted and mapped to the nearest supported value.	T2I Edit

Output	`size`		`aspect_ratio` + `resolution`
Flexible
Auto	`"auto"`	—	Model chooses optimal dimensions

1K 13 sizes

Output	`size`		`aspect_ratio` + `resolution`
1508 × 696	`"1508x696"`	or	`"19.5:9"` + `"1K"`
1183 × 887	`"1183x887"`	or	`"4:3"` + `"1K"`
1527 × 687	`"1527x687"`	or	`"20:9"` + `"1K"`
696 × 1507	`"696x1507"`	or	`"9:19.5"` + `"1K"`
1024 × 1024	`"1024x1024"`	or	`"1:1"` + `"1K"`
887 × 1182	`"887x1182"`	or	`"3:4"` + `"1K"`
687 × 1526	`"687x1526"`	or	`"9:20"` + `"1K"`
724 × 1448	`"724x1448"`	or	`"1:2"` + `"1K"`
1448 × 724	`"1448x724"`	or	`"2:1"` + `"1K"`
836 × 1254	`"836x1254"`	or	`"2:3"` + `"1K"`
1254 × 836	`"1254x836"`	or	`"3:2"` + `"1K"`
768 × 1365	`"768x1365"`	or	`"9:16"` + `"1K"`
1365 × 768	`"1365x768"`	or	`"16:9"` + `"1K"`

2K 13 sizes

Output	`size`		`aspect_ratio` + `resolution`
1392 × 3015	`"1392x3015"`	or	`"9:19.5"` + `"2K"`
1774 × 2365	`"1774x2365"`	or	`"3:4"` + `"2K"`
2365 × 1774	`"2365x1774"`	or	`"4:3"` + `"2K"`
1374 × 3053	`"1374x3053"`	or	`"9:20"` + `"2K"`
3053 × 1374	`"3053x1374"`	or	`"20:9"` + `"2K"`
1536 × 2731	`"1536x2731"`	or	`"9:16"` + `"2K"`
2731 × 1536	`"2731x1536"`	or	`"16:9"` + `"2K"`
2048 × 2048	`"2048x2048"`	or	`"1:1"` + `"2K"`
1448 × 2896	`"1448x2896"`	or	`"1:2"` + `"2K"`
2896 × 1448	`"2896x1448"`	or	`"2:1"` + `"2K"`
1672 × 2508	`"1672x2508"`	or	`"2:3"` + `"2K"`
2508 × 1672	`"2508x1672"`	or	`"3:2"` + `"2K"`
3014 × 1391	`"3014x1391"`	or	`"19.5:9"` + `"2K"`

How these parameters work

size

Exact pixel dimensions

"1920x1080"

aspect_ratio

Shape only, default scale

"16:9"

resolution

Scale tier, preserves shape

"1K"

Priority when combined

size › aspect_ratio + resolution › aspect_ratio › resolution

size is most specific and always wins. aspect_ratio and resolution control shape and scale independently.

How matching works

Shape matching – we pick the closest supported ratio. Ask for 7:1 on a model with 4:1 and 8:1, you get 8:1.

Scale matching – providers use different tier formats: K tiers (0.5K 1K 2K 4K) or megapixel tiers (0.25 1). If the exact tier isn't available, you get the nearest one.

Dimension clamping – if a model has pixel limits, we clamp dimensions to fit and keep the aspect ratio intact.

Media Inputs

Parameter	Type	Description	Modes
`image`	file	Required. Input image(s) to edit Supports PNG, JPEG, WebP.	T2I Edit

Output & Format

Parameter	Type	Description	Modes
`response_format` xai	string	Response format for generated images. If omitted, xAI returns URLs. `b64_json` `url` Default: `"url"`	T2I Edit
`output_format`	string	Output image format `png` `jpeg` `gif` `webp` `avif` Gateway converts to requested format if provider doesn't support it natively.	T2I Edit
`output_compression`	integer	Compression level for lossy formats (JPEG, WebP, AVIF)	T2I Edit
`n`	integer	Number of images to generate Default: `1` Gateway generates multiple images in parallel even if provider only supports 1.	T2I Edit

Additional Parameters

Provider-specific passthrough fields, available only when the request is routed to the listed provider.

Parameter	Type	Description	Modes
fal
`sync_mode`	boolean	If `True`, the media will be returned as a data URI and the output data won't be available in the request history.	T2I Edit
xai
`images`	array	Optional list of input images for multi-reference image editing. This field cannot be set together with `image`.	T2I Edit
`storage_options`	object	Options for storing generated output in the xAI Files API.	T2I Edit
`user`	["string", "null"]	Unique end-user identifier that can help xAI monitor and detect abuse.	T2I Edit

Parameter Normalization

How we handle parameters across different providers

Not every provider speaks the same language. When you send a parameter, we handle it in one of four ways depending on what the model supports:

Behavior	What happens	Example
`passthrough`	Sent as-is to the provider	style, quality
`renamed`	Same value, mapped to the field name the provider expects	prompt
`converted`	Transformed to the provider's native format	size
`emulated`	Works even if the provider has no concept of it	n, response_format

Parameters we don't recognize pass straight through to the upstream API, so provider-specific options still work.

Full API Reference Authentication, endpoints, and more

Grok Imagine Image Benchmarks

Grok Imagine Image currently holds an Elo rating of 1227 for text-to-image generation and an Elo of 1200 for image editing. It ranks 11th globally for image editing capabilities and 19th for general image generation performance.

Lumenfall Arena

#12

Image Editing

1201 Elo

Lumenfall Arena

#21

Text-to-Image

1237 Elo

Image Editing Landscape

Elo vs Cost

1 model without pricing omitted

Elo vs Speed

1 model waiting for enough speed data

Competition Results

Image Editing

Photorealism

View leaderboard

Grok Imagine Image edited result for Bald man challenge

Original image before Grok Imagine Image editing

Before After

Bald man challenge

15 models

Image Editing

Edit instruction

“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”

#12

Man and Car in California

13 models

Image Editing

Source

Edit instruction

“Make a photo of the man driving the car down the California coastline”

Grok Imagine Image edited result for Golden Hour Stroll

Before After

#11

Golden Hour Stroll

13 models

Image Editing

Edit instruction

“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”

#16

Outfit Transfer Challenge

16 models

Image Editing

Source

Edit instruction

“Use Image 1 as the base person. Dress them in the exact elaborate outfit from Image 2 (including all layers, accessories, jewelry, and shoes). Carefully adapt the clothing to the body shape and pose in Image 1 while maintaining realistic fabric behavior, correct proportions, and perfect lighting/shadow matching. Keep the person’s exact face, hair, and background completely unchanged.”

Grok Imagine Image edited result for Neutral Expression to Genuine Smile

Before After

Neutral Expression to Genuine Smile

14 models

Image Editing

Edit instruction

{
  "action": "image_edit",
  "reference": "uploaded neutral portrait",
  "change": "Warm genuine Duchenne smile: lips curved up, slight natural teeth, soft eye crinkles, subtle cheek raise",
  "details": "Realistic smiling skin (dimples if present, soft cheek shadows), slightly brighter eyes; keep exact eye shape/color/iris",
  "preserve_exact": "Face identity/structure, eyes/nose/lips/eyebrows, hair, skin texture/pores/freckles, makeup, clothing, head pose, background, lighting, shadows, framing",
  "no_changes": "No face shape change, no new features, no gaze shift, no hair/clothing/lighting/background edits",
  "style": "Ultra-photorealistic 8K portrait, sharp face focus, natural soft lighting, realistic skin glow"
}

3 attempts – showing best result

Grok Imagine Image edited result for Night Sky Transformation

Before After

Night Sky Transformation

16 models

Image Editing

Edit instruction

“Change the scene to night: a deep, dark sky with subtle, glistening stars visible behind the mountain.”

3 attempts – showing best result

Image Editing

Anime

View leaderboard

Studio Ghibli Anime Style

14 models

Image Editing

Source

Edit instruction

“Transform this photo into a Studio Ghibli–inspired illustration. Use soft pastel colors, hand-painted textures, gentle lighting, dreamy backgrounds, and a warm, nostalgic mood”

Image Editing

Portrait

View leaderboard

Before After

Bald man challenge

15 models

Image Editing

Edit instruction

“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”

Before After

Neutral Expression to Genuine Smile

14 models

Image Editing

Edit instruction

{
  "action": "image_edit",
  "reference": "uploaded neutral portrait",
  "change": "Warm genuine Duchenne smile: lips curved up, slight natural teeth, soft eye crinkles, subtle cheek raise",
  "details": "Realistic smiling skin (dimples if present, soft cheek shadows), slightly brighter eyes; keep exact eye shape/color/iris",
  "preserve_exact": "Face identity/structure, eyes/nose/lips/eyebrows, hair, skin texture/pores/freckles, makeup, clothing, head pose, background, lighting, shadows, framing",
  "no_changes": "No face shape change, no new features, no gaze shift, no hair/clothing/lighting/background edits",
  "style": "Ultra-photorealistic 8K portrait, sharp face focus, natural soft lighting, realistic skin glow"
}

3 attempts – showing best result

Uncategorized

Over-the-top cartoon caricature

13 models

Image Editing

Source

Edit instruction

“Create a caricature of me and my job. Make it exaggerated and humorous, incorporating my profession as a tv show anchor and my love for dogs and hockey.”

Top Matchups

See how Grok Imagine Image performs head-to-head against other AI models, ranked by community votes in blind comparisons.

vs Wan 2.6

Challenge: Golden Hour Stroll

0% W · 100% L

vs FLUX.2 [max]

Challenge: Studio Ghibli Anime Style

75% W · 25% L

vs Nano Banana

Challenge: Bald man challenge

0% W · 67% L · 33% T

vs GPT Image 1.5

Challenge: Vintage Cafe Logo

100% W · 0% L

vs GPT Image 1.5

Challenge: Modern Clean Menu

50% W · 50% L

Grok Imagine Image is best for

See all Use Cases

The model performs strongest in digital portraiture and text rendering, securing a 56.4% win rate in portraits and ranking 7th for text accuracy. Performance is more limited in anime and commercial branding categories, where win rates sit between 32% and 37.5%.

#5 of 15

Portrait

Edit · 57.9% win rate

Gallery

View all 8 images

Grok Imagine Image FAQ

How much does Grok Imagine Image cost?

Grok Imagine Image starts at $0.02 per image through Lumenfall. Pricing varies by provider. Lumenfall does not add any markup to provider pricing.

How do I use Grok Imagine Image via API?

You can use Grok Imagine Image through Lumenfall's OpenAI-compatible API. Send requests to the unified endpoint with model ID "grok-imagine-image". Code examples are available in Python, JavaScript, and cURL.

Which providers offer Grok Imagine Image?

Grok Imagine Image is available through xAI and fal.ai on Lumenfall. Lumenfall automatically routes requests to the best available provider.

What is the maximum resolution for Grok Imagine Image?

Grok Imagine Image supports images up to 2048x2048 resolution.

Overview

Grok Imagine Image is a high-fidelity text-to-image generation model developed by xAI. It is engineered to transform complex natural language prompts into visually striking, aesthetic imagery with a particular focus on realism and detailed composition. The model is distinctive for its adherence to user intent and its ability to render high-resolution outputs suitable for both creative exploration and commercial applications.

Strengths

Aesthetic Consistency: The model is tuned to prioritize visually appealing compositions, lighting, and textures, reducing the need for extensive “prompt engineering” to achieve professional-looking results.
Human Anatomy and Text Rendering: It demonstrates improved accuracy in rendering human features—such as hands and eyes—and can incorporate legible, coherent text within generated images more reliably than many first-generation diffusion models.
Prompt Adherence: The model excels at interpreting multi-layered instructions, accurately placing specific objects and following spatial relationships defined in the text description.
Processing Speed: Optimized for rapid inference, the model generates high-resolution images quickly, making it suitable for iterative design workflows.

Limitations

Style Bias: Because the model is optimized for “highly aesthetic” outputs, it may default to a polished or cinematic look even when a more raw or lo-fi aesthetic is requested.
Niche Concept Gaps: While strong on general concepts, the model may occasionally struggle with highly technical or obscure domain-specific imagery where training data density is lower.
Image Editing Constraints: While capable of image-to-image tasks, it may lack the granular “in-painting” controls found in specialized tools dedicated solely to localized image manipulation.

Technical Background

Grok Imagine Image is built upon a concentrated diffusion architecture designed by xAI, leveraging massive datasets to bridge the gap between semantic understanding and visual synthesis. Its training approach emphasizes “alignment” between the latent visual space and conversational language patterns, allowing it to understand prompts that are phrased naturally rather than as a string of keywords.

Best For

This model is ideal for creating marketing collateral, concept art, and high-quality social media assets where visual impact is the primary goal. It is also well-suited for rapid prototyping in UI/UX design and architectural visualization. Grok Imagine Image is available through Lumenfall’s unified API and playground, allowing developers to integrate high-end image generation into their applications with minimal overhead.

Try Grok Imagine Image in Playground

Generate images with custom prompts — no API key needed.

Open Playground