Grok Imagine Image 2.0 makes 10 images per request; Sume's n cap
xAI says grok-imagine-image-2.0 returns up to 10 images per request at $0.04 each. Sume's n is 1-10 per request, with a lower per-model cap on x-ai/grok-image.

xAI's guide says grok-imagine-image-2.0 can generate up to 10 images per request at a flat per-image price, and its models page lists that id at $0.04 per image. Sume's n accepts 1 to 10 per request, but each model has its own ceiling in the catalog, and the repo default for catalog max_images is 4. Read the n range for x-ai/grok-image from GET /v1/images/models before you ask for ten.
The two sides
Prices and limits are from xAI's docs and the Sume docs; the Sume figures are static repo list rates, so check the live pricing line.
| Item | xAI | Sume |
|---|---|---|
| Images per request | Up to 10 (grok-imagine-image-2.0) | n 1-10 request cap, per-model cap from catalog |
| Price shown | $0.04 for -2.0, $0.05 for -quality, $0.02 for grok-imagine-image | x-ai/grok-image list $0.02, billed list x 1.25 |
| Edit sources | Up to 5 source images | input_references, up to 10 on models that support edits (read the descriptor) |
Ten images in dollars
Sume bills cost_usd x n. At the repo's list rate of $0.02, billed at x 1.25, one image is $0.025 before any rounding. For ten images that is $0.25 unrounded, or $0.30 if each image were rounded up to the cent, against 10 x $0.04 = $0.40 on xAI's -2.0 line. The ids are different products, and the Sume docs do not say which xAI id x-ai/grok-image uses, so this is a price gap, not a like-for-like comparison.
What to do when the cap is below ten
If the catalog cap is 4, ten images take three requests: 4 + 4 + 2. A request above the cap is rejected by the catalog, so read the descriptor first. Slow, large-n jobs can also exceed the 30-second wait and come back as a 202 job.
curl "https://api.sume.com/v1/images/models/x-ai/grok-image/endpoints" \
-H "Authorization: Bearer $SUME_API_KEY"Where to read it
In the response, find supported_parameters.n for the cap and pricing[0].cost_usd for the billed line.
Sources
Related posts
More in Developers
- H3 Max job: poll or webhook? A signed Python receiver
Poll GET /v1/jobs/:id/status for one clip, use a signed webhook for batches. Python receiver that refuses an empty secret and checks the sume-v1 signature.
- Lip sync for singing: fal can turn off guidance, Sume has no switch
fal's H3 Max Lip Sync is transcription-guided by default and can be switched off for singing or processed audio. Sume's documented body has no such field.
- HEAD-check reference image URLs before an image edit call (Python)
OpenAI caps edit images at 50 MB and Ideogram at 25 MB. A Python pre-flight that checks size, type and https before you send references to Sume.
- Hold AI Shorts for human approval: poll the run, then publish
A Python poll loop for a Sume Format run that stops at a human yes or no before any upload, so a reviewer can reject sameness first.
Written by Sume