Grok Imagine image prices: $0.02, $0.04, $0.05 and Sume's one row

xAI prices its image models on three rungs: grok-imagine-image at $0.02 per image, grok-imagine-image-2.0 at $0.04, and grok-imagine-image-quality at $0.05. Sume's image catalog has a single Grok row, x-ai/grok-image. Sume's docs do not say which of the three xAI ids that row maps to, so do not assume one. Read the row's live price before you budget.

The xAI ladder

All three figures come from xAI's model page, read on 2026-10-05. The 1,000-image column is plain multiplication: price times 1,000.

• The xAI image guide (read 2026-10-05) says grok-imagine-image-2.0 returns up to 10 images per request at a flat per-image price.
• Edits take up to 5 source images, and xAI bills both the input images and the generated output images. Budget an edit at more than a plain generation.

What Sume lists

Sume bills provider list price times 1.25, and the Image API reports the amount in usage.cost as billed USD. Sume's static rate table has the Grok row at $0.02 before margin. The arithmetic: $0.02 x 1.25 = $0.025 before any rounding. The live per-endpoint pricing array is the authoritative number, so quote that, not this arithmetic.

Read the live price

One request shows the billed line for the Grok row. The pricing array carries cost_usd for each billable line, and the docs say you pay cost_usd x n.

import asyncio, os, httpx

async def main():
    url = "https://api.sume.com/v1/images/models/x-ai/grok-image/endpoints"
    headers = {"Authorization": f"Bearer {os.environ['SUME_API_KEY']}"}
    async with httpx.AsyncClient(timeout=30) as c:
        r = await c.get(url, headers=headers)
    r.raise_for_status()
    for ep in r.json()["endpoints"]:
        print(ep["pricing"])

asyncio.run(main())

How to choose

If you need to match an xAI rung exactly, call xAI directly and pick the id on the table above. If you want one billing line across image families, call the Grok row on Sume and compare its endpoints price with the other catalog rows through the same call. See the Sume Image API docs for the response shape, and the 10-image request comparison for the n cap.

Before you budget

Pick the rung by the job, then price it from the same source you will bill against. If your team already sends traffic to xAI, the ladder is the whole story. If you route through Sume, ask the endpoints call for the line and multiply by your image count, since the docs say you pay cost_usd x n. Re-read both pages before a large run: prices and model ids change, and this post is a dated snapshot, read 2026-10-05.

• Edits on xAI bill the input images as well as the output, so a 3-image edit costs more than 3 plain generations.
• Failed requests on Sume are not billed, per the Image API docs.