Firefly Video Model 1
The first Firefly Video Model began a limited public beta in October 2024 and reached general availability in April 2025. It generates footage from text or image inputs. Adobe’s published API still identifies its video route as video1_standard.
Start and end images guide the shot
The product guide permits a first frame, a last frame, or both, alongside a text description. These images guide how the clip begins and ends. Resolution and aspect ratio are separate choices. A supplied final frame does not specify every intermediate movement.Firefly Video generation and frame controls
The product label does not always show the version
Practical guidance. The editing interface calls the choice Firefly Video, while the API uses a versioned header and Adobe’s newer training disclosure names Video Model 2. Keep the exact route in a workflow record. Do not assume a product label and a numbered API model necessarily refer to the same deployed revision.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Launch date | |
| Access | API, Provider product |
| Model identifiers | Firefly Videovideo1_standard |
| API route | video1_standard on /v3/videos/generate.Adobe’s published Firefly API schema |
| Product defaults | The Firefly Video editor guide gives five seconds and 24 fps as defaults.Firefly Video generation and frame controls |
Related models
- Firefly Sound Effects Model
Adobe · Firefly Audio
Sound-effects model that uses text to describe sounds and optional voice input to control timing and energy.
Launch date Provider product - Firefly Video Model 2
Adobe · Firefly Video
Second-generation Firefly video model disclosed in August 2026, without a documented public API identifier.
Launch date not confirmedAccess not confirmedProvider product