AI Video Generation Cost Calculator

List prices are per generated second — but nobody ships the first take. Set how many generations one finished clip really takes, and this calculator shows what you actually pay per usable clip and per month.

Per generation
–
Per finished clip
–
after all takes
Monthly budget
–

All models compared at your settings

Model — planRate/sPer generationPer finished clipMonthly

Each model uses its default plan at the closest available resolution to your selection. Click a row in the calculator above by changing the model — the highlighted row is your current pick.

Full published rate table

Per-second API list rates, verified 2026-10-12. Channels differ per model — aggregators frequently price the same model 2–3× apart, which is exactly why the channel is named on every row.

Veo 3.1 · Google · up to 8s clips

Google's flagship video model via the Gemini API. Audio is included in every tier.

Plan1080p720p4K
Lite$0.080/s$0.050/s—
Fast$0.12/s$0.10/s$0.30/s
Standard$0.40/s$0.40/s$0.60/s

Source: Gemini API pricing (Google), rates as published Sep–Oct 2026. Rate source ↗ · Open the Veo 3.1 calculator →

Gemini Omni 1.1 Flash · Google · up to 10s clips

Conversational video generation and editing; GA since Aug 27, 2026. The 360p draft tier is built for cheap iteration.

Plan1080p360p4K720p
Standard (GA)$0.15/s$0.030/s$0.30/s$0.10/s

Source: Gemini API pricing (Google), GA rate ladder Aug 2026. Rate source ↗ · Open the Gemini Omni 1.1 Flash calculator →

Kling 3.0 · Kuaishou · up to 15s clips

Multi-shot 'AI Director' model with native audio and a Motion Control variant. Official developer API rates.

Plan1080p720p4K
No native audio$0.11/s$0.084/s—
With native audio$0.17/s$0.13/s—
4K——$0.42/s

Source: Kling official developer pricing, as published Sep 2026. Rate source ↗ · Open the Kling 3.0 calculator →

Seedance 2.5 · ByteDance (BytePlus) · up to 30s clips

Up to 30-second clips and 50 reference assets. Official BytePlus ModelArk rates, token-based billing converted at 16:9. BytePlus does not sell in the US; reseller prices differ.

Plan1080p480p720p
BytePlus ModelArk (official)$0.57/s$0.10/s$0.23/s

Source: BytePlus ModelArk model pricing, checked Oct 6, 2026. Rate source ↗ · Open the Seedance 2.5 calculator →

Seedance 2.0 · ByteDance · up to 15s clips

Previous generation; widest aggregator availability. Rates below are per-channel — the same model varies 2–3x between providers.

Plan1080p480p720p
Standard via Replicate$0.45/s$0.080/s$0.18/s
Mini via ModelsLab—$0.044/s$0.092/s

Source: Replicate / ModelsLab listed rates, checked Sep 27, 2026. Rate source ↗ · Open the Seedance 2.0 calculator →

Vidu Q4 Preview · Shengshu Technology · up to 16s clips

Launched Oct 7, 2026. Up to 15 reference images and 3 reference audio clips, 4K/10-bit. Image- and reference-to-video only. Audio does not change the price.

Plan1080p2K4K540p720p
List price$0.12/s$0.19/s$0.39/s$0.045/s$0.095/s
Launch promo −30% (I2V/R2V, thru Nov 30, 2026)$0.084/s$0.13/s$0.27/s$0.032/s$0.067/s

Source: Vidu official price list (credits at $0.005/credit), Oct 2026. Rate source ↗ · Open the Vidu Q4 Preview calculator →

Grok Imagine Video 1.5 · xAI · up to 15s clips

xAI direct API rates per output second. Aggregators resell the same model far below these rates — channel choice matters more than model choice here.

Plan1080p480p720p
xAI API (direct)$0.25/s$0.080/s$0.14/s

Source: xAI API pricing, as documented Oct 2026. Rate source ↗ · Open the Grok Imagine Video 1.5 calculator →

Grok Imagine Video · xAI · up to 15s clips

Base Imagine video model on the xAI API; 480p/720p only.

Plan480p720p
xAI API (direct)$0.050/s$0.070/s

Source: xAI API pricing, as documented Oct 2026. Rate source ↗ · Open the Grok Imagine Video calculator →

Wan 3.0 · Alibaba · up to 30s clips

30-second multimodal clips (even documents as input) via Alibaba Model Studio international list rates. A 30% launch promo ran through Sep 24, 2026 and has ended; list rates shown.

Plan1080p480p720p
Standard$0.20/s$0.050/s$0.10/s
Prime (faster)$0.28/s$0.068/s$0.14/s

Source: Alibaba Model Studio international list rates, Aug 2026. Rate source ↗ · Open the Wan 3.0 calculator →

How to read these prices

Per finished clip is the only number that matters. A $0.40/s model where you keep 1 take in 2 beats a $0.10/s model where you keep 1 in 8. Set “generations per finished clip” honestly — for character work, 4–6 is typical; for simple B-roll, 2–3.

Channel beats model. Seedance 2.5 at 720p is $0.23/s on BytePlus and $0.47/s on fal.ai for the same weights. Grok Imagine 1.5 is $0.14/s at 720p on xAI's own API and a fraction of that on some aggregators. Always price the door, not just the model.

Promos expire. Vidu Q4's launch pricing runs through November 30, 2026, and Wan 3.0's launch discount already ended in September. Every rate on this page carries its verification date; if a vendor has gone quiet for a month, re-check before budgeting a big run.