- Sora 2 is the cheapest frontier option at $0.10 per second of 720p video — $0.05 on the batch tier. Sora 2 Pro runs $0.30 (720p) to $0.70 (1080p).
- Veo 3.1 Standard costs $0.40 per second at 720p/1080p and $0.60 at 4K, four times Sora 2. Veo 3.1 Fast drops to $0.10, and Veo 3.1 Lite to $0.05.
- Kling 2.5 Turbo Pro is $0.07 per second and Wan 2.5 is $0.05 per second through fal.
- Google has shut Veo 3 down, on 30 June 2026, and Imagen 4 on 17 August 2026, so most Veo 3 pricing pages on the web now quote a model that no longer runs.
- A 30-second clip costs $1.50 on Veo 3.1 Lite, $3.00 on Sora 2, and $12.00 on Veo 3.1 Standard.
AI video generation cost is quoted per second of output rather than per request, which makes it unusually easy to compare across providers and unusually easy to underestimate. A rate that looks trivial at $0.40 becomes $24 for a single one-minute clip, and a workflow that regenerates a shot ten times before it is usable multiplies that again. Every figure below comes from the provider own pricing page, checked on 18 August 2026.
AI video pricing: cost per second, compared
| Model | Resolution | Price / second | Source |
|---|---|---|---|
| Wan 2.5 | — | $0.05 | fal |
| Veo 3.1 Lite | 720p | $0.05 | |
| Kling 2.5 Turbo Pro | — | $0.07 | fal |
| Veo 3.1 Lite | 1080p | $0.08 | |
| Sora 2 | 720p | $0.10 | OpenAI |
| Veo 3.1 Fast | 720p | $0.10 | |
| Veo 3.1 Fast | 1080p | $0.12 | |
| Veo 3.1 Fast | 4K | $0.30 | |
| Sora 2 Pro | 720p | $0.30 | OpenAI |
| Veo 3.1 Standard | 720p / 1080p | $0.40 | |
| Sora 2 Pro | 1024p | $0.50 | OpenAI |
| Veo 3.1 Standard | 4K | $0.60 | |
| Sora 2 Pro | 1080p | $0.70 | OpenAI |
The spread from cheapest to dearest is 14x: $0.05 on Veo 3.1 Lite or Wan 2.5 against $0.70 on Sora 2 Pro at 1080p. That gap is far wider than the quality gap most projects would notice, which is the single most useful thing to know before picking a default.
What a real clip actually costs
| Model | 10 seconds | 30 seconds | 60 seconds |
|---|---|---|---|
| Veo 3.1 Lite (720p) | $0.50 | $1.50 | $3.00 |
| Kling 2.5 Turbo Pro | $0.70 | $2.10 | $4.20 |
| Sora 2 (720p) | $1.00 | $3.00 | $6.00 |
| Veo 3.1 Fast (1080p) | $1.20 | $3.60 | $7.20 |
| Veo 3.1 Standard (720p/1080p) | $4.00 | $12.00 | $24.00 |
| Sora 2 Pro (1080p) | $7.00 | $21.00 | $42.00 |
These are the cost of one successful generation. Video models are not one-shot tools: a usable shot commonly takes three to ten attempts as you adjust the prompt, seed or reference image. Budget the table above multiplied by your real retry count, and the difference between a $0.05 and a $0.40 model stops being academic.
Sora 2 pricing
OpenAI charges $0.10 per second for sora-2 at 720p, in either portrait (720×1280) or landscape (1280×720). The Pro tier is priced by resolution: $0.30 at 720p, $0.50 at 1024p and $0.70 at 1080p.
The detail most comparisons miss is the batch tier, which halves everything. sora-2 falls to $0.05 per second and sora-2-pro to $0.15 / $0.25 / $0.35. If your generation is not interactive, such as overnight renders, bulk asset production or anything queued, batch turns Sora 2 into the cheapest frontier video model available. Interactive work pays double for the latency.
Veo 3.1 pricing, and the Veo 3 trap
The current Google line is Veo 3.1, in three tiers. Standard is $0.40 per second at 720p and 1080p, rising to $0.60 at 4K. Fast is $0.10 (720p), $0.12 (1080p) and $0.30 (4K). Lite is $0.05 (720p) and $0.08 (1080p). All tiers include audio by default.
Google shut Veo 3 down on 30 June 2026. This matters because Veo 3 pricing is still one of the most-searched terms in this category, and much of the content answering it describes a model that is being retired. If you are choosing now, price Veo 3.1; if you have Veo 3 in production, the migration is not optional.
The three-tier structure is the useful part. An eight-fold price difference between Lite and Standard at the same nominal resolution means the tier choice, not the vendor choice, is usually where the money goes.
Kling and the cheaper alternatives
Through fal, Kling 2.5 Turbo Pro is $0.07 per second and Wan 2.5 is $0.05 per second. Rates for models routed through an aggregator can differ from the vendor own console, so treat these as that aggregator rate rather than a universal one. At these levels they undercut every first-party frontier tier except Veo 3.1 Lite.
For high-volume, lower-stakes work such as social clips, background footage and concept passes, the $0.05 to $0.07 band does the job at a seventh of Veo 3.1 Standard.
AI image generation pricing
Image pricing does not use the same unit, which makes direct comparison harder.
| Model | Price | Unit |
|---|---|---|
| Imagen 4 Fast | $0.02 | per image |
| Imagen 4 Standard | $0.04 | per image |
| Imagen 4 Ultra | $0.06 | per image |
| gpt-image-1 | $10 in / $40 out | per 1M tokens |
| gpt-image-1-mini | $2.50 in / $8 out | per 1M tokens |
OpenAI bills images as tokens rather than per picture, so the cost of a single image depends on its size and detail instead of being a flat rate. That is harder to forecast, but cheaper for small outputs. Batch halves those token rates as well.
Google shut Imagen 4 down on 17 August 2026, recommending Gemini 2.5 Flash Image instead. As with Veo 3, a lot of published Imagen 4 pricing now describes a retiring model.
How to choose without overpaying
Start at the bottom of the range, not the top. Most projects default to the flagship tier and never test whether Lite or Fast is good enough for the shot they actually need. At an eight-fold price difference inside a single vendor, that test is worth running before you scale.
Use batch wherever the work is not interactive. Halving the rate is a larger saving than any vendor switch, and it costs nothing but latency.
Count retries, not generations. The honest unit of cost is seconds of finished video, which at a typical five-attempt hit rate is five times the sticker price.
Check the deprecation notice before you build. Two of the most-quoted models in this category are on published shutdown schedules.
If your workload also involves text models, the same per-unit discipline applies there. The AI API cost calculator works out monthly spend per model, and the AI models database carries specifications and pricing for every major model we track. For running generation locally instead, the VRAM calculator sizes the GPU you would need.
Per-model pages, kept current from our models database: Sora 2 · Sora 2 Pro · Veo 3.1 · Kling 2.5 Turbo Pro · Wan 2.5.
Frequently asked questions
How much does Sora 2 cost per second?
Sora 2 costs $0.10 per second of 720p video on the standard tier and $0.05 per second on the batch tier. Sora 2 Pro is $0.30 per second at 720p, $0.50 at 1024p and $0.70 at 1080p, halving to $0.15 / $0.25 / $0.35 on batch.
Is Veo cheaper than Sora?
It depends entirely on the tier. Veo 3.1 Lite at $0.05 per second is cheaper than Sora 2 at $0.10, and Veo 3.1 Fast matches it at $0.10. But Veo 3.1 Standard at $0.40 per second is four times the price of Sora 2. Comparing vendors without naming the tier is meaningless in this category.
What does a one-minute AI video cost?
For one successful generation: $3.00 on Veo 3.1 Lite, $4.20 on Kling 2.5 Turbo Pro, $6.00 on Sora 2, $24.00 on Veo 3.1 Standard and $42.00 on Sora 2 Pro at 1080p. Multiply by your retry rate, because a usable shot rarely lands on the first attempt.
Is Veo 3 still available?
Google shut Veo 3 down on 30 June 2026 and directs new work to Veo 3.1. Much of the Veo 3 pricing content published online still describes the retiring model, so check the tier and the version before budgeting against it.
Why is AI video priced per second instead of per video?
Generation cost scales almost linearly with output length, because the model produces frames sequentially. Per-second billing therefore reflects the provider actual compute more honestly than a flat per-video fee would, and it is why a short clip is cheap while a one-minute render is not.
What is the cheapest AI video generator API?
Among the models with published rates, Veo 3.1 Lite at 720p and Wan 2.5 both sit at $0.05 per second, with Kling 2.5 Turbo Pro just above at $0.07. Sora 2 on the batch tier also reaches $0.05, which makes it the cheapest frontier-quality option for non-interactive work.
Prices are the published list rates from each provider, checked on 18 August 2026 against the OpenAI API pricing page, the Google Gemini API pricing page and the fal pricing page. Aggregator rates can differ from a vendor own console, and all providers change pricing without notice.
