Veo vs Runway

The short answer

Both figures are list price for one 8 second clip. Eight seconds is a length both pages actually sell. Neither figure is a lotgen quote. Neither is a measured keep rate. There is no footage on this page.

A lower printed list price is not a verdict about which clip you keep.

What the keyword is asking

veo vs runway asks what Google’s Veo list costs next to Runway’s developer-portal credits, before anyone runs a lot.

A same-prompt quality comparison is a different job. It needs the same prompt, a real run, and a keep rate. That is not this page. There is no lotgen run footage to put beside these prices.

The two cells share a duration both routes sell: 8 seconds. Gemini’s parameter table also sells 4 and 6 seconds. At 1080p and 4k, that table says the length must be 8 seconds. Runway’s Gen-4.5 request sells any whole number of seconds from 2 to 10. Gemini does not sell a 5 second Veo clip, so there is no 5 second Veo figure here.

Someone reading the query is usually deciding which bill they are looking at. The dollars below are what each provider publishes for the row named in the table. They are not a quote for a specific run.

One list-price row each

Veo, from Gemini

One worked row, from Gemini API pricing, retrieved 2026-09-22.

Paid tier, Veo 3.1 Standard, video with audio (the default row): $0.40 per second at 720p and at 1080p. At 8 seconds, list price is $0.40 x 8 = $3.20. The same row prints $0.60 per second at 4k. Free tier for this model is marked not available. Google says you are charged only if the video is successfully generated.

Model id on that page: veo-3.1-generate-preview. The guide marks Veo 3.1 as Preview.

The full Standard, Fast, and Lite table lives on Veo pricing explained.

Duration and aspect, same day, from Generate videos with Veo 3.1: durationSeconds is “4”, “6”, or “8”, and must be “8” for extension, reference images, or 1080p and 4k. Aspect is 16:9 (default) or 9:16.

Runway, from the developer portal

One worked row, from Runway’s API pricing page, retrieved 2026-09-22.

• Credits bought in the developer portal are $0.01 per credit. Sales tax may apply.
• gen4.5 is 12 credits per second.
• 8 seconds is 96 credits.
• List price: 96 x $0.01 = $0.96.

That page does not split Gen-4.5 by resolution, so there is no separate 720p or 1080p dollar rate to quote. The $0.96 is the base rate. ProRes, PNG sequences, 10-bit, and HDR add a per-second credit surcharge on Gen-4.5. Those surcharges are not in the $0.96.

This row is Runway’s own gen4.5 model id. The portal prices other model ids on the same page. Gen-4 Turbo, on that page, is 5 credits per second. It is not the worked row.

The same portal table also lists veo3.1 (audio) at 40 credits per second and veo3.1 (no audio) at 20 credits per second. At $0.01 per credit, those lines are $0.40 per second and $0.20 per second. Eight seconds of veo3.1 (audio) is 320 credits, or $3.20. Gemini’s Standard cell for 8 seconds at 720p or 1080p is also $3.20. Same arithmetic is two invoices: one is Runway credits for a model id on Runway’s portal, and one is Gemini’s paid tier. The worked Runway cell stays Gen-4.5. The veo3.1 credit lines are not a second comparison, and they are not Gemini’s bill.

What the official docs actually limit

No quality scores. Only limits with a provider doc behind them. Checked on 2026-09-22.

Veo 3.1 on Gemini

Sources: Gemini API pricing and the Veo 3.1 guide. The parameter column below is the guide’s column for Veo 3.1 and Veo 3.1 Fast, which is the column that covers Standard.

• One generation: 4, 6, or 8 seconds. The length must be 8 seconds at 1080p, at 4k, with reference images, or when extending. The guide’s opening line also describes Veo 3.1 as a model for 8 second videos at 720p, 1080p, or 4k. The worked cell uses 8 seconds, which both sentences allow.
• Resolution: 720p (default), 1080p (8 seconds only), 4k (8 seconds only). Extension is 720p only.
• Aspect: 16:9 (default) or 9:16. No 1:1 in that list.
• Audio: the feature table says natively generated audio is always on. The price row is video with audio. The guide says you can prompt sound effects, ambient noise, and dialogue.
• Videos per request: 1.
• Extension, Veo 3.1 and Veo 3.1 Fast, not Lite: extend a previous Veo video by 7 seconds, up to 20 times. The guide says the combined output can run up to 148 seconds. Input for an extension is a Veo-generated video up to 141 seconds, at 720p, at 9:16 or 16:9. Videos are stored for 2 days.
• personGeneration: text-to-video and extension are allow_all only. Image-to-video, interpolation, and reference images are allow_adult only. In EU, UK, CH, and MENA locations, the guide says allow_adult is the only allowed value.
• The guide lists request latency as a minimum of 11 seconds and a maximum of 6 minutes during peak hours.
• Status: Preview. Generated videos are watermarked with SynthID.
• You are charged only if the video is successfully generated. The guide also says a blocked audio or safety result is not charged.

Runway Gen-4.5

Sources: API pricing and the API reference, text-to-video and image-to-video for model gen4.5.

• Duration: 2 to 10 seconds on both text-to-video and image-to-video. Max length on these calls is 10 seconds.
• Ratio, text-to-video: 1280:720 or 720:1280. 720:1280 is 9:16. 1:1 is not in this list.
• Ratio, image-to-video: 1280:720, 720:1280, 1104:832, 960:960, 832:1104, 1584:672. 960:960 is 1:1. 720:1280 is 9:16.
• Audio: neither Gen-4.5 request lists an audio field. The outputFormat description says a PNG sequence returns a zip of PNG frames plus a separate wav artifact when the output has audio. A missing audio field does not mean “the clip is silent”, and the wav clause does not mean “Gen-4.5 generates native audio”.
• Prompt text: required, max length 1000.
• The 12 credits per second rate is the base model rate. HDR and professional output formats add credits on top. They are documented on the pricing page, and they are not in the $0.96 example.

Runway’s veo3.1 request, on the same API reference, is a different model id. Its duration is 4, 6, or 8 seconds. It has an audio boolean, default true, and the pricing page says audio inclusion affects that price. Its ratios are 1280:720, 720:1280, 1080:1920, and 1920:1080. That request is the portal credit line above. It is not the Gen-4.5 worked row, and it is not the Gemini parameter table.

Where the other pages live

One line each on what the neighbouring pages own:

• Veo pricing explained owns the full Standard, Fast, and Lite table. This page is one Standard cell.
• AI video generation cost owns usable-shot math. List price for one clip is not that math.
• Kling vs Veo cost is a different pair.
• The cost calculator is a separate tool. It is not the source of the list prices above.
• Runway vs Kling owns Runway versus Kling, including the Kling row.
• AI video API pricing is the twelve-key grid, for a reader who needs the wider set of keys.
• Best AI video generators is the overview of how the keys differ.

What this page will not decide

This page will not tell you which clip looks better, what share of takes you keep, or what a same-prompt pair costs after a real run. A same-prompt quality comparison is a different job. There is no footage here, and no stills are invented to stand in for one.

It will not rank a winner. The $3.20 cell and the $0.96 cell are list prices for the rows named above, at 8 seconds. They are not a quote for a take you have run.

It will not repeat the Fast and Lite grid, the Kling row, the twelve-key grid, or the usable-shot math. Those live on the pages named above.

It will not turn Runway’s app subscription into a dollar per second. The worked Runway cell is the developer portal.

Request access

lotgen is where a Veo key and a Runway key can be quoted before you commit to a lot. The seat is not the clip. The labs still bill the clip. The figures on this page stay list price. A quote is a later step, on a real request.

Request an invitationSeats and pricing

Team aside