Grok Imagine Image 2.0 API on WaveSpeedAI: Parameters, Pricing and How It Compares to Earlier Grok Image Endpoints

xAI’s Grok Imagine Image 2.0 is available on WaveSpeedAI as two endpoints, live since 2026-08-16: x-ai/grok-imagine-image-v2.0/text-to-image at a flat $0.05 per image and x-ai/grok-imagine-image-v2.0/edit at a flat $0.06 per image. Both take a resolution of 1k or 2k and a quality of low or medium, and the price does not change with either. Text-to-image offers 13 aspect ratios; edit takes exactly one input image. Below are the full parameter tables, what xAI announced, how these endpoints compare to the three earlier Grok image endpoints in the catalog, and where each fits.

What xAI announced

xAI published Imagine Image 2.0 on 2026-08-07 as the new Quality Mode on grok.com/imagine and in its iOS and Android apps. The announcement emphasizes instruction following and typography: the model plans text layout and composition before rendering, so text-dense outputs such as posters and infographics keep their structure. It also lists editing tools for the consumer surface: a magic wand for region-targeted edits, segmentation, background removal with transparent export, multi-reference editing with up to five input images, and smart resize across nine aspect ratios. xAI states that the model ranked second on both the Arena text-to-image and image-edit leaderboards as of the launch date.

On the API side, xAI’s model documentation lists the model as grok-imagine-image-2.0, and a separate migration notice schedules the older grok-imagine-image-quality model for retirement on 2026-11-02, with requests redirecting to 2.0 at low quality after that date.

Two caveats for developers. The magic wand, segmentation and background removal are app tools, not API parameters, and none of them appear in the WaveSpeedAI schema. And the five-image multi-reference editing is a consumer-app feature; the WaveSpeedAI edit endpoint accepts one image per request.

Parameters

Text-to-image

Parameter	Required	Values	Default
prompt	yes	text description of the image	-
aspect_ratio	no	1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 2:1, 1:2, 19.5:9, 9:19.5, 20:9, 9:20	1:1
resolution	no	1k, 2k	2k
quality	no	low, medium	medium

One image is generated per request; there is no num_images field on this endpoint. The 19.5:9 and 9:19.5 ratios match modern phone screens, and 20:9 / 9:20 cover the taller Android formats, which is useful for full-bleed app store and social assets.

Edit

Parameter	Required	Values	Default
images	yes	array with at most one image URL; jpg, jpeg, png, webp	-
prompt	yes	edit instruction	-
resolution	no	1k, 2k	2k
quality	no	low, medium	medium

The edit endpoint has no aspect_ratio field; output follows the source image. The listed price already includes the input-image fee, so there is no surcharge for the source.

Pricing

Both live formulas are total_price = base_price, so the cost is fixed per request regardless of parameters:

Endpoint	model_uuid	Price per image	Varies with resolution/quality
Text-to-image	x-ai/grok-imagine-image-v2.0/text-to-image	$0.05	no
Edit	x-ai/grok-imagine-image-v2.0/edit	$0.06	no

A thousand 2K medium-quality generations is $50; a thousand edits is $60. Because the tiers are free, the sensible default is 2k and medium unless you are throughput-bound, in which case 1k returns faster.

How 2.0 compares to the earlier Grok image endpoints

The catalog carries three Grok image generations. Here is how they line up, using each endpoint’s live schema and formula:

Endpoint family	model_uuid prefix	Text-to-image price	Edit price	Resolution tiers	Batch (num_images)	Edit inputs
Grok Imagine Image (original)	x-ai/grok-imagine-image/	$0.022	$0.025	none	up to 4	1 image
Grok Imagine Image Quality	x-ai/grok-imagine-image-quality/	$0.06 at 1K, $0.08 at 2K	$0.07 at 1K, $0.09 at 2K, plus $0.01 per extra input image	1k, 2k	up to 4	up to 3 images
Grok Imagine Image 2.0	x-ai/grok-imagine-image-v2.0/	$0.05	$0.06	1k, 2k	1 per request	1 image

Three things follow from the table.

First, 2.0 is cheaper than the Quality endpoints at every combination: $0.05 versus $0.06 or $0.08 for generation, $0.06 versus $0.07 or $0.09 for edits. Since xAI is retiring the Quality model in November, 2.0 is the natural target for anyone on grok-imagine-image-quality today.

Second, the original x-ai/grok-imagine-image/ endpoints remain the cheapest option in the family at $0.022 per image, and they still support batches of up to four. If you need volume over fidelity, for thumbnails, placeholder art or first-pass concepting, they are not superseded on price.

Third, 2.0 dropped two conveniences: batch generation and multi-image edit input. A workload that relied on num_images: 4 needs four requests now, and an edit workflow that combined three references needs to be restructured around a single source image with the change described in the prompt.

Typical use cases

• Text-heavy marketing assets: posters, banners and infographic-style images, where xAI’s typography planning claim is the differentiator over the earlier generations. Use 2k and a wide or tall ratio.
• Product photography variants: generate a base image, then use the edit endpoint to change color, background or lighting per SKU. At $0.06 per edit a 20-variant set is $1.20.
• Mobile-first creative: the 9:19.5 and 9:20 ratios produce assets that fill a phone screen without cropping.
• Style exploration at fixed cost: because the price does not change with tier, run 1k low while iterating prompts and switch to 2k medium only for the final render, without any change in budget per image.

Calling the endpoints

Every WaveSpeedAI model uses the same request pattern: POST https://api.wavespeed.ai/api/v3/<model_uuid> with a JSON body, then poll GET /api/v3/predictions/<id>/result until the job completes. An edit request looks like this:

curl -X POST https://api.wavespeed.ai/api/v3/x-ai/grok-imagine-image-v2.0/edit \
  -H "Authorization: Bearer $WAVESPEED_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"images": ["https://example.com/sneaker-white.png"], "prompt": "Change the sneaker color to forest green, keep the background and lighting unchanged", "resolution": "2k", "quality": "medium"}'

Model pages with the live schema, playground and pricing:

• Grok Imagine Image 2.0 text-to-image
• Grok Imagine Image 2.0 edit

Related

The video side of the Grok Imagine family has its own landing pages: Grok Imagine Video and Grok Imagine Video 1.5. For the earlier image endpoints, see the original write-ups of Grok Imagine Image text-to-image and Grok Imagine Image edit on WaveSpeedAI.