# Video models — video_attributes fields and live queries

Source: https://ofox.io/docs/api/videos/models

All video models on [`POST /v1/videos`](https://ofox.io/docs/api/videos/create) share the same request structure; they differ in the resolutions, durations, aspect ratios, and generation modes they accept. Each model's capabilities are described by the `video_attributes` object returned by the [Models API](https://ofox.io/docs/api/openai/models). This page explains that object's fields and how to query it.

For each model's parameter values, see the family pages: [Seedance](https://ofox.io/docs/develop/models/seedance), [Wan & HappyHorse](https://ofox.io/docs/develop/models/wan), [Hailuo](https://ofox.io/docs/develop/models/hailuo).

## The `video_attributes` object

Every video model in the [Models API](https://ofox.io/docs/api/openai/models) response carries a `video_attributes` object, which is the authoritative description of the parameters that model accepts.

```json
{
  "modes": ["t2v", "i2v", "v2v"],
  "resolutions": ["480p", "720p", "1080p"],
  "default_resolution": "720p",
  "min_duration_seconds": 4,
  "max_duration_seconds": 30,
  "supports_audio": true,
  "aspect_ratios": ["21:9", "16:9", "4:3", "1:1", "3:4", "9:16", "adaptive"]
}
```

| Field | Type | Description |
|------|------|------|
| `modes` | string[] | Generation modes the model supports: `t2v` text-to-video, `i2v` image-to-video, `v2v` video as reference. You do not pass a mode in the request; it is inferred from the input fields. See [Mode detection](https://ofox.io/docs/api/videos/create#modes) |
| `resolutions` | string[] | Accepted `resolution` values; values outside the list are rejected |
| `default_resolution` | string | Applied when the request omits both `resolution` and `size` |
| `min_duration_seconds` | number | Lower bound of `duration`, in seconds |
| `max_duration_seconds` | number | Upper bound of `duration`, in seconds |
| `supports_audio` | boolean | Whether `generate_audio: true` produces an audio track |
| `aspect_ratios` | string[] | Accepted `aspect_ratio` values; `adaptive` keeps the input image's ratio |

`video_attributes` is the source of truth for a single model's `resolution`, `duration`, and `aspect_ratio`. The parameter ranges listed in [Create video task](https://ofox.io/docs/api/videos/create) are the union across all models, so a value valid there may still be rejected by a specific model.

## Query the live list

The Models API is a public endpoint and requires no API key. Filter on `supported_endpoints` to keep only video models:

**cURL**

```bash filename="Terminal"
curl -s https://api.ofox.io/v1/models \
  | jq -r '.data[] | select(.supported_endpoints[]? == "/v1/videos") | .id'
```

**Python**

```python filename="video_models.py" showLineNumbers
import requests

models = requests.get("https://api.ofox.io/v1/models").json()["data"]

for m in models:
    if "/v1/videos" in m.get("supported_endpoints", []):
        va = m["video_attributes"]
        print(m["id"], va["modes"], va["resolutions"], va["max_duration_seconds"])
```

**TypeScript**

```typescript filename="video-models.ts" showLineNumbers
const resp = await fetch('https://api.ofox.io/v1/models')
const { data } = await resp.json()

for (const m of data) {
  if (!m.supported_endpoints?.includes('/v1/videos')) continue
  const va = m.video_attributes
  console.log(m.id, va.modes, va.resolutions, va.max_duration_seconds)
}
```

New models go live independently of docs releases. When you need the latest list, query this endpoint directly.

## Next

  - **Seedance family**
  - **Wan & HappyHorse family**
  - **Hailuo family**
  - **Create video task**
  - **Provider routing**
  - **Video API overview**