Veo 3.1 Fast 4K at $2.40 for 8 seconds vs Omni 4K on Sume

If you want a 4K AI video clip with sound and are comparing prices, there are two different questions: what Google charges for Veo 3.1 Fast, and what you can actually call through Sume. They have different answers, and the second one is narrower.

What Google lists

Google's Gemini API pricing page lists Veo 3.1 Fast at $0.10 per second at 720p, $0.12 at 1080p and $0.30 at 4K, with audio included (read 2026-10-05). An 8-second 4K clip is therefore $2.40 at list.

What Sume lists

The Sume video catalog does not include Veo 3.1. The /v1/videos contract lists Veo 3.1 as a family that is not in the v1 catalog because there is no provider entry, builder or pricing constant for it. The catalog id that generates 4K natively with audio is gemini-omni-flash-1.1 (minimax-h3 can also bill 4K, but as an upscale of 768p output). It accepts 3 to 10 seconds at 360p, 720p, 1080p or 4K, in 16:9 or 9:16, with native synced audio. Its fal list is $0.30 per second at 4K, so Sume bills $0.375 per second, rounded up per job.

How to decide

Ask the catalog before you plan: GET /v1/video-router/models is the source of truth for ids and rates.

• The $0.60 difference is the Sume multiple. You are paying for one wallet, one API shape and one job envelope across models.
• If you need Veo specifically, call Google directly; do not send model: "veo-3.1" to Sume, which returns 404 model_not_found.
• If 4K just needs to be sharp, check whether 1080p ($1.50 for 8 seconds on Omni) is enough for the screen it lands on.