Gemini Omni 1.1 Flash vs Veo 3.1 vs Wan 3.0 Pricing
The Short Answer
Veo 3.1 Lite if you want the lowest per-second price. Wan 3.0 if you want one long continuous take. Gemini Omni 1.1 Flash if the clip will be revised more than once.
Google shipped Gemini Omni 1.1 Flash on August 27, 2026, two days after Alibaba launched Wan 3.0 on August 25, 2026. The two launches landed in the same week and they are optimising for opposite things — Google for iterative editing control, Alibaba for length and raw cost.
The Price Table
Per output second, as of August 28, 2026:
| Gemini Omni 1.1 Flash | Veo 3.1 Lite | Veo 3.1 Fast | Veo 3.1 Standard | Wan 3.0 | |
|---|---|---|---|---|---|
| Vendor | Alibaba | ||||
| 360p / 480p | $0.03 | — | — | — | $0.05 |
| 720p | $0.10 | $0.05 | $0.10 | ~$0.40 | $0.10 |
| 1080p | $0.15 | $0.08 | $0.12 | ~$0.40 | $0.20 |
| 4K | $0.30 | — | $0.30 | — | — |
| Native clip length | 10s | ~8s | ~8s | ~8s | 30s |
| Max sequence | 40s (extension) | — | — | — | 30s |
| Open weights | No | No | No | No | No |
Veo 3.1 Standard rates vary by surface and region; treat the ~$0.40 figure as indicative and confirm on Google’s official pricing page before budgeting production volume.
Cost Per Finished Minute — the Number That Actually Matters
Per-second rates flatter models that make you generate more attempts. A more honest metric is cost per usable minute, which folds in the re-roll rate.
Assume a 60-second 1080p deliverable and a 3× re-roll factor (three generations per usable shot — conservative for text-to-video work):
| Model | Raw 60s | With 3× re-rolls | With draft-first workflow |
|---|---|---|---|
| Veo 3.1 Lite | $4.80 | $14.40 | n/a (no draft tier) |
| Gemini Omni 1.1 Flash | $9.00 | $27.00 | ~$12.60 |
| Veo 3.1 Fast | $7.20 | $21.60 | n/a |
| Wan 3.0 | $12.00 | $36.00 | ~$21.00 (480p drafts) |
The draft-first column is why Omni 1.1’s $0.03 tier matters. Iterate composition at 360p, render final once at 1080p, and the effective cost drops by more than half — landing close to Veo 3.1 Lite while keeping the editing controls Lite does not have.
Completion criterion: you know your own re-roll rate from logs before you pick a model. If you do not, measure it on 20 shots before committing budget.
Control Features Compared
| Feature | Omni 1.1 Flash | Veo 3.1 | Wan 3.0 |
|---|---|---|---|
| Conversational scene extension | ✅ 10s increments to 40s | ❌ | ❌ |
| Start + end frame locking | ✅ | Partial | ❌ |
| External footage as style ref | ✅ up to 3s | ❌ | ✅ via Omni-Reference |
| Document / spreadsheet input | ❌ | ❌ | ✅ Omni-Reference |
| Audio in same pass | ✅ | ✅ (Fast/Standard) | ✅ |
| Cheap draft tier | ✅ 360p | ❌ | 480p at $0.05 |
Wan 3.0’s differentiator is input breadth — its Omni-Reference system accepts documents, spreadsheets, presentations and public web pages as source material and generates video from them. That is a marketing and training-content play, not a creative-generation one.
Gemini Omni 1.1 Flash’s differentiator is output revision — the ability to keep extending and adjusting one sequence coherently.
Which One to Pick
Pick Veo 3.1 Lite when the job is high-volume, short, one-shot clips and per-second cost dominates. Nothing else in this comparison beats $0.08 per 1080p second.
Pick Gemini Omni 1.1 Flash when shots get revised — ad variants, explainers, anything with a review cycle. The draft tier plus 40-second coherent extension is a real pipeline, and it is the only model here where iteration is priced differently from delivery.
Pick Wan 3.0 when you need a single continuous 30-second take with audio in one pass, or when your source material is documents rather than prompts. Note the caveat: Alibaba shipped Wan 2.1 and 2.2 with open weights and did not do so for 3.0, so the self-hosting escape hatch that made earlier Wan releases attractive is closed.
None of the three can be run locally. If that is a hard requirement, this entire comparison is the wrong shortlist — you want the older open Wan checkpoints or an open video model, and you will trade quality for it.