Adobe Firefly’s Generate video feature does not hand you one video model. It hands you a Model dropdown, and Adobe’s own Firefly Video model sits in it beside Google’s Veo 3.1.
That makes this a settings decision rather than a vendor argument. Same prompt field, different option lists, different bill, and in one case different audio.
One dropdown, two owners
Adobe’s generation guide states the default in one sentence: the model used by default will be Firefly Video, but if you are generating a video for the first time, the default will be Veo 3.1.
The two entries are not the same kind of thing. Firefly Video is Adobe’s own model, described in Adobe’s documentation as the way to create commercially safe video clips using simple text descriptions. Veo 3.1 and Veo 3.1 Fast are Google’s, listed on Adobe’s partner-model page under the heading Generate videos using Veo 2, Veo 3.1, and Veo 3.1 Fast.
Frame rate is the one setting that ignores the choice. Adobe states that the frames per second for all generated videos is set to a default of 24 FPS across Adobe and partner models. Everything else is model-specific, which Adobe concedes in a note: the settings available for video generation may differ depending on the video model you select.
Firefly Video vs Veo 3.1: the settings each one exposes
Both columns describe what appears after you pick the model inside Firefly, and both come from Adobe’s own documentation.
| Setting | Firefly Video (Adobe’s model) | Veo 3.1 (partner model) |
|---|---|---|
| Published by | Adobe | |
| Resolution | 540p, 720p or 1080p | 720p or 1080p |
| Aspect ratio | Widescreen (16:9), Vertical (9:16), Square (1:1) | Widescreen (16:9), Vertical (9:16) |
| Duration | 5 seconds | 4, 6 or 8 seconds |
| Frames per second | 24 FPS | 24 FPS |
| Audio option | Not offered in the generation step | An Audio toggle for synchronized sounds or music |
| Credits | 100 per second at 1080p, 50 at 720p, 20 at 540p | Listed as varies |
Three rows carry the decision. Duration is fixed at five seconds on Adobe’s model while Veo 3.1 offers 4, 6 or 8. Square framing exists only on Adobe’s model, and audio only on Veo.
What each vendor publishes outside Firefly
Google’s Veo page states that the model generates outputs in 1080p and 4K, and that Veo 3 lets you add sound effects, ambient noise, and even dialogue to your creations, generating all audio natively. A Google blog post adds native vertical 9:16 output.
Adobe’s list is narrower. For its own model the resolution options are 540p, 720p or 1080p, with 1080p recommended for the best results. In the credits documentation, 4K appears only for partner models such as Ray3.14 and for Premiere’s Generative Extend, because the resolution table for Generate Video itself stops at 1080p.
Native audio is the row that decides most picks
Adobe’s sentence about Veo is specific: Veo 3.1 and Veo 3.1 Fast offer the ability to generate audio along with your video, and enabling the Audio option generates synchronized sounds or music for the generated video.
No equivalent sentence exists for Firefly Video. There, sound arrives after the clip does, through a Generate sound effects action that adds sound effects using text prompts or your own voice recording. For Adobe’s own model, native audio is not published in either direction: Adobe documents a post-generation audio workflow, not a native audio capability.
Google publishes the limit of its own audio too. Its Veo page says that creating videos with natural and consistent spoken audio, particularly for shorter speech segments, remains an area of active development.
How the credits land on each model
Firefly Video is metered by the second. Adobe’s generative credits FAQ lists 1080p at 24 FPS at 100 credits per second, 720p at 50 and 540p at 20, which makes a five-second clip 500, 250 or 100 credits. The Firefly Video cost guide works that arithmetic through, and the free plan boundaries decide how many clips a month cost nothing.
Veo is a separate line. Adobe marks partner-model consumption as varies and explains that the number of credits used will depend on the model selection, output, and file size. There is no rate to multiply, so a Veo clip cannot be budgeted the way Adobe’s own model can.
Example: one prompt, two model selections
Switch only the Model dropdown and the two blocks below diverge on their own. The prompt is Adobe’s, published with the Firefly Video Model announcement.
cinematic realistic detailed shot of a cute robot holding up a red glowing heart. The lighting is gorgeous and sun-kissed, with dappled lighting on the robot's face and a strong sunny backlight, shallow depth of field. film grain. shot on 35mm film.Case 1: Firefly Video for a silent first pass
Model: Firefly Video
Resolution: 540p
Aspect ratio: Widescreen (16:9)
Duration: 5 seconds
Frames per second: 24 FPS
Audio: none in this step
Credits: 100Adobe’s defaults put every generated clip at 24 FPS and five seconds, so the dial you are really turning is resolution. At 540p the clip costs 20 credits per second, a fifth of the 1080p price. The model exposes no audio option in the generation step, so sound is a second job: Generate sound effects. The step-by-step guide shows where each control sits.
Case 2: Veo 3.1 on the same prompt with audio on
Model: Veo 3.1
Resolution: 1080p
Aspect ratio: Widescreen (16:9)
Duration: 8 seconds
Frames per second: 24 FPS
Audio: on
Credits: variesTwo things change beyond the model name. Duration becomes a choice of 4, 6 or 8 seconds, and the Audio toggle appears. The credits also stop being predictable, because Adobe’s only statement about partner models is varies.
What neither vendor page settles
Google does not publish a per-clip duration for Veo on the pages this site could reach; the Vertex AI documentation did not open at all. Adobe’s resolution list for its own model stops at 1080p while Google publishes 1080p and 4K for Veo, so those 4K rows belong either to Veo outside Firefly or to other partner models inside it. And for Firefly Video there is no native-audio statement in either direction, only a documented post-generation sound workflow.
Frequently asked questions
Which model does Firefly use by default? Firefly Video, except on your first generation, when Adobe makes the default Veo 3.1.
Which one has native audio? Veo 3.1 and Veo 3.1 Fast, per Adobe’s note that enabling the Audio option generates synchronized sounds or music. Adobe publishes no equivalent sentence for Firefly Video.
Why does duration look different on each model? Because the models expose different settings. Adobe says the settings available for video generation may differ depending on the video model you select.
Can I run the same prompt on both? Yes, and that is what the dropdown is for. Swap it and everything else stays put. The Firefly Video prompt guide covers the five-slot structure that makes those runs worth comparing.