Grok Imagine Video 1.5 Lite vs Sume: clip cost by length

xAI's docs list Grok Imagine Video 1.5 Lite at a flat price per second of output. Sume does not list the Lite model; it lists grok-imagine-video-1.5, an image-to-video model at 480p and 720p. The table compares the two for one clip and for a hundred clips at each length.

Worked example: 100 10-second Grok Imagine 1.5 Lite clips vs Sume 720p

xAI's docs list Grok Imagine Video 1.5 Lite at $0.020 per second of output, so 100 10-second clips come to $20.00 direct. Sume does not list the Lite model. It lists grok-imagine-video-1.5, an image-to-video model at 480p and 720p, which bills $12.50 for the same 100 clips.

xAI's figure is from its model page, read 2026-10-10; the page gives one per-second rate and does not split it by resolution. Sume's figures come from the Video Router docs and API pricing.

What does a 10-second clip cost on each route?

The Sume rows are every catalog model that accepts 10 seconds at 720p, cheapest first, with the Grok row included for reference. Sume's Grok row takes a source image; the others also take text-only prompts.

How do I call it?

A text-to-video request goes to POST /v1/videos. This one uses the cheapest text-capable row; Sume's Grok model needs a frame_images entry instead of a text-only prompt.

import os
import time

import requests

H = {"Authorization": f"Bearer {os.environ['SUME_API_KEY']}"}
job = requests.post(
    "https://api.sume.com/v1/videos",
    headers=H,
    json={
        "model": "minimax-h3",
        "prompt": "A paper boat drifts down a rain-soaked street at dusk",
        "duration": 10,
        "resolution": "768p",
        "aspect_ratio": "16:9",
    },
).json()
while True:
    s = requests.get(job["polling_url"], headers=H).json()
    if s["status"] in ("completed", "failed"):
        break
    time.sleep(10)
print(s["status"], s.get("unsigned_urls"))

What this comparison leaves out

The two Grok routes are different products.

• Lite and the model Sume lists are separate xAI releases; do not read the table as the same model at two prices.
• xAI's page does not state the resolution its $0.020 rate applies to.
• Output quality is not compared.

Every clip length in one table

This page replaces 24 posts that asked the same question with different numbers. Each row is one of them, with the answer it gave when it was published.