What is Sora2 Pro True 1080p?
Sora2 Pro True 1080p is an AI video generation model available through UniAll AI. It supports both **text-to-video** and **reference image-to-video** workflows, with selectable output resolutions including `720p`, `1080p`, and `true_1080p`.
Use the public model id exactly as:
```text sora2-pro-true-1080p ```
The model is designed for short-form, high-quality video generation such as product reveals, cinematic social clips, ad creatives, e-commerce visuals, storyboards, and vertical or horizontal marketing videos.
Who should use this API?
Sora2 Pro True 1080p is a good fit for:
- Developers building AI video generation features into apps or internal tools
- Marketing teams producing short video assets at scale
- E-commerce teams creating product showcase clips
- Agencies generating draft creatives, storyboards, and ad variants
- Platforms that need async video generation with predictable parameters
Because the API is asynchronous, it is better suited for production workflows than blocking request-response generation.
Supported generation modes
The API supports two main modes:
| Mode | Use case | Required inputs | |---|---|---| | `text_to_video` | Generate a video from a prompt | `prompt` | | `reference_image_to_video` | Animate or build from a reference image | `prompt`, `image_url` |
Supported aspect ratios:
- `16:9` for horizontal video
- `9:16` for vertical short-form video
Supported durations:
- `4`, `8`, `12`, `16`, or `20` seconds
Supported resolutions:
- `720p`
- `1080p`
- `true_1080p`
Only one video is generated per request: `video_count: 1`.
API endpoint
```http POST /v1/videos/generations ```
This is an async video generation endpoint. Submit a generation request, then poll or handle the task result according to your UniAll AI integration flow.
Text-to-video request example
```bash curl -X POST "https://api.uniall.ai/v1/videos/generations" \ -H "Authorization: Bearer YOUR_UNIALL_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "sora2-pro-true-1080p", "generation_mode": "text_to_video", "prompt": "A cinematic product reveal of a matte black smartwatch rotating on a glass pedestal, soft studio lighting, shallow depth of field, smooth camera movement.", "duration": 8, "aspect_ratio": "16:9", "resolution": "true_1080p", "video_count": 1 }' ```
Reference image-to-video request example
```bash curl -X POST "https://api.uniall.ai/v1/videos/generations" \ -H "Authorization: Bearer YOUR_UNIALL_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "sora2-pro-true-1080p", "generation_mode": "reference_image_to_video", "prompt": "Animate this product photo into a premium commercial shot. Add a slow dolly-in camera move, subtle reflections, and elegant studio lighting.", "image_url": "https://example.com/reference.png", "duration": 4, "aspect_ratio": "16:9", "resolution": "720p", "video_count": 1 }' ```
Required and optional parameters
| Parameter | Type | Required | Notes | |---|---:|---:|---| | `model` | string | Yes | Use `sora2-pro-true-1080p` | | `generation_mode` | string | Yes | `text_to_video` or `reference_image_to_video` | | `prompt` | string | Yes | Describe scene, camera, motion, style, and subject | | `image_url` | string | For image-to-video | Accepts PNG, JPEG, or WebP reference image URL | | `duration` | integer | Yes | `4`, `8`, `12`, `16`, or `20` seconds | | `aspect_ratio` | string | Yes | `16:9` or `9:16` | | `resolution` | string | Yes | `720p`, `1080p`, or `true_1080p` | | `video_count` | integer | Optional | Fixed at `1` |
Prompting tips for better video results
A strong prompt usually includes:
1. **Subject**: what appears in the video 2. **Action or motion**: what changes over time 3. **Camera direction**: dolly-in, pan, orbit, handheld, locked-off 4. **Lighting and style**: cinematic, studio, natural daylight, macro, commercial 5. **Output intent**: ad creative, product demo, social clip, storyboard
Example:
```text A premium skincare bottle on wet stone, slow macro dolly-in, water droplets sliding down the surface, soft morning light, luxury commercial style, realistic reflections, 9:16 vertical video. ```
Pricing angle
Sora2 Pro True 1080p is billed per generated second. Pricing varies by selected resolution:
| Resolution | User price | |---|---:| | `720p` | $0.204 / second | | `1080p` | $0.34 / second | | `true_1080p` | $0.476 / second |
For example, an 8-second `true_1080p` generation is estimated at:
```text 8 × $0.476 = $3.808 ```
Use `720p` for faster drafts and cost control, then switch to `true_1080p` for final assets that need higher visual quality. UniAll AI supports refund-on-failure behavior for failed generations, while automatic retry is not enabled by default for this model.
Practical implementation notes
- Use async task handling rather than expecting the video to return immediately.
- Validate `duration`, `resolution`, and `aspect_ratio` before sending requests.
- Keep uploaded reference images accessible by URL when using `reference_image_to_video`.
- Store request metadata so failed or pending jobs can be traced.
- For batch workflows, queue requests because each request returns one video.
When to choose true_1080p
Choose `true_1080p` when the final video will be used in campaigns, landing pages, client delivery, or high-quality product showcases. For early ideation, `720p` or standard `1080p` can reduce cost while keeping the same creative workflow.
常见问题
What is the model id for Sora2 Pro True 1080p API?
Use the public model id `sora2-pro-true-1080p` when calling the UniAll AI video generation endpoint.
Does Sora2 Pro True 1080p support image-to-video?
Yes. Use `generation_mode: reference_image_to_video` and provide both `prompt` and `image_url`. The reference image can be PNG, JPEG, or WebP.
How is Sora2 Pro True 1080p priced?
It is billed per generated second. Current user pricing is $0.204/second for 720p, $0.34/second for 1080p, and $0.476/second for true_1080p.