Grok Imagine 1.5 Lite at 2 cents vs 1.5 at 8 cents: what Sume bills
xAI lists Lite at $0.020 and 1.5 at $0.080 a second. Sume carries only the 1.5 row, image-to-video, 4-15 s, with its own list rate. Which line applies.

Sume bills one Grok video row, grok-imagine-video-1.5, at 1.25 times a list rate of $0.01 per second, which is $0.0125 per second; it has no Lite row. xAI's models page lists two lines: grok-imagine-video-1.5-lite at $0.020 per second and grok-imagine-video-1.5 at $0.080 per second (xAI models and pricing, read 2026-10-10).
The two numbers for 1.5 do not match, and this post explains why you should not mix them: Sume's rate comes from the catalog entry in the repo, not from xAI's page, so use the catalog for any Sume budget.
What xAI's pages say
The xAI Lite model page (read 2026-10-10) lists modalities as text and image to video, a price of $0.020 per second, a rate limit of 10 requests per second, two regions and Batch API support. It leaves durations, resolutions and aspect ratios unspecified, so this post does not repeat the 1 to 15 second range or the 1080p tier that other sites quote.
The video guide says video generation is priced per second with both duration and resolution affecting the total, and that duration is configurable up to 15 seconds.
What Sume's Grok row is
In Sume's catalog the row is image-to-video only, 480p or 720p, 4 to 15 seconds, no aspect_ratio, no audio toggle, no end frame and no reference videos or audio. The first input must be a still. The repo's list rate is 10,000 USD micros per second, billed at the house 1.25 margin, so the reserve at submit is list times 1.25 per second (Video generation docs).
Sume's published video wire reports the same facts at runtime: read supported_durations, supported_resolutions and pricing_skus from GET /v1/videos/models rather than trusting a blog table.
| Line | Per second | Where |
|---|---|---|
| grok-imagine-video-1.5-lite | $0.020 | xAI list; no Sume row |
| grok-imagine-video-1.5 | $0.080 | xAI list |
| grok-imagine-video-1.5 | $0.01 list, $0.0125 billed | Sume catalog (list x 1.25) |
Worked arithmetic for a 6-second clip
A 6-second Grok clip on Sume reserves 6 x $0.0125 = $0.075. The xAI Lite line would give 6 x $0.020 = $0.12 and the xAI 1.5 line 6 x $0.080 = $0.48 for the same length, but those are xAI's own billing; do not expect them on a Sume invoice.
If you need Lite-class pricing on Sume with text input, the choice is a different family, not a Lite id. Wan 3.0 starts at $0.05 list per second at 480p ($0.0625 billed), and Gemini Omni Flash 1.1 starts at $0.03 list at 360p ($0.0375 billed). Both take a text prompt.
Decision rule
Use the Sume Grok row when you already have a still and want a short clip. Use another row when the job needs text-only input, a last frame, references or an aspect ratio.
- Still in hand, 4 to 15 seconds:
grok-imagine-video-1.5on Sume. - Text only: Wan 3.0 or Omni Flash.
- Need Lite id exactly: call xAI directly; Sume does not list it.
Sources
Related posts
More in Pricing
- Haiku 5.5 with US-only inference: 1.1x on both price tiers
Anthropic's pricing page applies the 1.1x US-only multiplier to Haiku 5.5's over-100k prices too. A worked agent-loop table and where Sume's own cap sits.
- Ideogram V3 is $0.075 at every quality on Sume; 4.5 low is half
Sume bills Ideogram V3 a flat $0.075 per image whatever quality you send, while Ideogram 4.5 at low quality is $0.0375. Medium is the same on both rows.
- Nano Banana 2.1 fal price fell to $0.040; Sume still bills $0.10
Nano Banana 2.1 launched cheaper upstream: fal read $0.040 at 1K on 2026-10-10, Sume's catalog bills $0.10. The gap by tier and how to budget.
- Rehearse an image prompt for 2.5 cents, then render the keeper
Test an image prompt on a 2.5 cent Sume model before paying for Nano Banana Pro. Eight drafts plus one final cost 0.3875 USD versus 1.6875 for nine finals.
Written by Sume