Grok Imagine Image 2.0 returns up to 10 images: n on Sume

xAI's Imagine API docs say grok-imagine-image-2.0 can return up to 10 images per request and prices them flat per image regardless of prompt length. Sume lists the Grok image model as x-ai/grok-image; n is capped by the catalog row, not by xAI's figure, so check supported_parameters.n.max before you ask for 10.

xAI's limits are from its image generation guide, read 2026-10-03. Sume's rules are from the Image API docs.

What xAI documents

The guide names grok-imagine-image-2.0 for image generation and editing and grok-imagine-video-1.5 for video. It lets you edit one image from a URL or base64 data URI, or combine up to 5 source images in one request, and says generated media is subject to content policy review and is not used for training.

Read the Sume ceiling

The catalog lists each row's n and input_references range. Run this and find the x-ai/grok-image line.

import os
import requests

key = os.environ["SUME_API_KEY"]
url = "https://api.sume.com/v1/images/models"
resp = requests.get(url, headers={"Authorization": f"Bearer {key}"}, timeout=30)
resp.raise_for_status()
for model in resp.json()["data"]:
    params = model["supported_parameters"]
    refs = params.get("input_references", {}).get("max", 0)
    n_max = params.get("n", {}).get("max", 1)
    print(model["id"], "refs:", refs, "n:", n_max)

Limits

A request that exceeds the model's range is rejected with 400 unsupported_parameter, not trimmed. Billing follows the docs: completed generations are billed in full and failed ones are not, so a batch that fails costs nothing, but a batch that completes costs cost_usd times n.