---
title: MiniMax H3 | Runware Docs
url: https://runware.ai/docs/models/minimax-h3
description: Multimodal video generation with native synced audio, multi-reference consistency, and continuation workflows
---
# MiniMax H3

![MiniMax H3](https://assets.runware.ai/covers/minimax-h3.jpg)

MiniMax H3 is a multimodal video generation model that supports text-to-video, first-frame and keyframe-guided generation, multi-reference conditioning, and audio-video continuation in a single workflow. It accepts text together with images, videos, and audio references to keep subjects, voice, motion, and scene identity more consistent across shots, while generating synchronized sound natively rather than as a separate dubbing pass. It is well suited to cinematic multi-shot generation, reference-driven character performance, instruction-based video editing, and continuation workflows that extend an existing clip or audio segment into a seamless new video.

- **Status**: live
- **ID**: `minimax:h3@0`
- Try in Playground
- **Creator**: MiniMax
- **Release Date**: July 30, 2026
- **Zero Data Retention**: Yes
- **Capabilities**: Text to Video, Image to Video, Video to Video, Audio to Video, Edit, Checkpoint, Audio

## Pricing

$0.08 from, per second · ~5m 8s per request

**Rates**

- 768p · per second: `$0.08`
- 2K · per second: `$0.13`
- each image beyond the first fiveper input image · beyond 5: `$0.04`
- uploaded video · per second: `$0.13`

**4 example generations**

- [text-to-video](https://runware.ai/docs/models/minimax-h3/examples#example-avalanche-patrol-team-portrait) · 1344×768 · 6s: `$0.48` (~2m 27s)
- [first-frame](https://runware.ai/docs/models/minimax-h3/examples#example-indoor-wasabi-farm-b-roll) · 768p · 8s: `$0.64` (~2m 54s)
- [first-last-frame](https://runware.ai/docs/models/minimax-h3/examples#example-underwater-scooter-colorway-reveal-film) · 1440p · 10s: `$1.30` (~7m 27s)
- [reference-to-video](https://runware.ai/docs/models/minimax-h3/examples#example-wastewater-lab-public-health-briefing) · 768p · 12s: `$1.74` (~7m 22s)

times vary, measured end to end · [How pricing works](https://runware.ai/docs/models-api/pricing)

## Compatibility & Validation

Inside `inputs`, `frameImages` cannot be used with `referenceImages`, `referenceVideos`, or `referenceAudios`.

---

When `inputs.frameImages` is provided, `width/height` cannot be used.

---

When `inputs.referenceAudios` is provided, `inputs.referenceImages` or `inputs.referenceVideos` is required.

---

`resolution` cannot be used with `width/height`.

---

When `resolution` is provided, `inputs.frameImages`, `inputs.referenceImages`, or `inputs.referenceVideos` is required.

---

`width` and `height` must be used together.

---

The following dimension combinations are supported:

| Configuration | Dimensions |
| --- | --- |
| `768p (~16:9)` | `1344x768` |
| `768p (4:3)` | `1024x768` |
| `768p (1:1)` | `768x768` |
| `768p (3:4)` | `768x1024` |
| `768p (~9:16)` | `768x1344` |
| `768p (~21:9)` | `1536x672` |
| `1440p (16:9)` | `2560x1440` |
| `1440p (4:3)` | `1920x1440` |
| `1440p (1:1)` | `1440x1440` |
| `1440p (3:4)` | `1440x1920` |
| `1440p (9:16)` | `1440x2560` |
| `1440p (~21:9)` | `2944x1248` |

## Request Parameters

**API Options**

Platform-level options for task execution and delivery.

### [taskType](https://runware.ai/docs/models/minimax-h3#request-tasktype)

- **Type**: `string`
- **Required**: true
- **Value**: `videoInference`

Identifier for the type of task being performed.

### [taskUUID](https://runware.ai/docs/models/minimax-h3#request-taskuuid)

- **Type**: `string`
- **Required**: true
- **Format**: `UUID v4`

UUID v4 identifier for tracking tasks and matching async responses. Must be unique per task.

### [outputType](https://runware.ai/docs/models/minimax-h3#request-outputtype)

- **Type**: `string`
- **Default**: `URL`

Video output type.

**Allowed values**: `URL`

### [outputFormat](https://runware.ai/docs/models/minimax-h3#request-outputformat)

- **Type**: `string`
- **Default**: `MP4`

Specifies the file format of the generated output. The available values depend on the task type and the specific model's capabilities.

- \`MP4\`: Widely supported video container (H.264), recommended for general use.
- \`WEBM\`: Optimized for web delivery.
- \`MOV\`: QuickTime format, common in professional workflows (Apple ecosystem).

**Allowed values**: `MP4` `WEBM` `MOV`

### [outputQuality](https://runware.ai/docs/models/minimax-h3#request-outputquality)

- **Type**: `integer`
- **Min**: `20`
- **Max**: `99`
- **Default**: `95`

Compression quality of the output. Higher values preserve quality but increase file size.

### [webhookURL](https://runware.ai/docs/models/minimax-h3#request-webhookurl)

- **Type**: `string`
- **Format**: `uri`

Webhook URL that receives JSON responses via HTTP POST when generation tasks complete. For batch requests with multiple results, each completed item triggers a separate webhook call as it becomes available.

**Learn more** (1 resource):

- [Webhooks](https://runware.ai/docs/models-api/webhooks) (platform)

### [deliveryMethod](https://runware.ai/docs/models/minimax-h3#request-deliverymethod)

- **Type**: `string`
- **Default**: `async`

Determines how the API delivers task results.

**Allowed values**:

- `sync` Returns results directly in the API response. Over HTTP, a task still running after 120 seconds continues as an async task and comes back with status 'processing'. Poll for its results using getResponse.
- `async` Returns an immediate acknowledgment with the task UUID. Poll for results using getResponse.

**Learn more** (1 resource):

- [Task Polling](https://runware.ai/docs/models-api/task-polling) (platform)

### [uploadEndpoint](https://runware.ai/docs/models/minimax-h3#request-uploadendpoint)

- **Type**: `string`
- **Format**: `uri`

Specifies a URL where the generated content will be automatically uploaded using the HTTP PUT method. The raw binary data of the media file is sent directly as the request body. For secure uploads to cloud storage, use presigned URLs that include temporary authentication credentials.

**Common use cases:**

- **Cloud storage**: Upload directly to S3 buckets, Google Cloud Storage, or Azure Blob Storage using presigned URLs.
- **CDN integration**: Upload to content delivery networks for immediate distribution.

```text
// S3 presigned URL for secure upload
https://your-bucket.s3.amazonaws.com/generated/content.mp4?X-Amz-Signature=abc123&X-Amz-Expires=3600

// Google Cloud Storage presigned URL
https://storage.googleapis.com/your-bucket/content.jpg?X-Goog-Signature=xyz789

// Custom storage endpoint
https://storage.example.com/uploads/generated-image.jpg
```

The content data will be sent as the request body to the specified URL when generation is complete.

### [safety](https://runware.ai/docs/models/minimax-h3#request-safety)

- **Path**: `safety.checkContent`
- **Type**: `object (2 properties)`

Content safety checking configuration for video generation.

#### [checkContent](https://runware.ai/docs/models/minimax-h3#request-safety-checkcontent)

- **Path**: `safety.checkContent`
- **Type**: `boolean`

Enable or disable content safety checking. Increases total generation time and adds a small charge on top of the generation cost.

#### [mode](https://runware.ai/docs/models/minimax-h3#request-safety-mode)

- **Path**: `safety.mode`
- **Type**: `string`
- **Default**: `fast`

Safety checking mode for video generation.

**Allowed values**:

- `fast` Checks key frames.
- `full` Checks all frames.

### [ttl](https://runware.ai/docs/models/minimax-h3#request-ttl)

- **Type**: `integer`
- **Min**: `60`

Time-to-live (TTL) in seconds for generated content. Only applies when `outputType` is `URL`.

### [includeCost](https://runware.ai/docs/models/minimax-h3#request-includecost)

- **Type**: `boolean`

Include task cost in the response.

### [numberResults](https://runware.ai/docs/models/minimax-h3#request-numberresults)

- **Type**: `integer`
- **Min**: `1`
- **Max**: `4`
- **Default**: `1`

Number of results to generate. Each result uses a different seed, producing variations of the same parameters.

**Inputs**

Input resources for the task (images, audio, etc). These must be nested inside the \`inputs\` object.

### [referenceImages](https://runware.ai/docs/models/minimax-h3#request-inputs-referenceimages)

- **Path**: `inputs.referenceImages`
- **Type**: `array of strings`
- **Min items**: `1`
- **Max items**: `9`
- **Size (each)**: `256 px to 5760 px`
- **Aspect ratio (each)**: `1:2.5 to 2.5:1`
- **File size (each)**: `up to 30 MB`

List of reference images (UUID, URL, Data URI, or Base64).

### [frameImages](https://runware.ai/docs/models/minimax-h3#request-inputs-frameimages)

- **Path**: `inputs.frameImages`
- **Type**: `array of strings or objects`
- **Min items**: `1`
- **Max items**: `2`
- **Size (each)**: `256 px to 5760 px`
- **Aspect ratio (each)**: `1:2.5 to 2.5:1`
- **File size (each)**: `up to 30 MB`

An array of frame-specific image inputs to guide video generation. Each item can be either a plain image input (UUID, URL, Data URI, or Base64) or an object that pairs an image with a target position in the video.

The `frameImages` parameter allows you to constrain specific frames within the video sequence, ensuring that particular visual content appears at designated points. Position can be specified using `frame` (named positions or frame indices) or `timestamp` (seconds), depending on model support. This is different from `referenceImages`, which provide overall visual guidance without constraining specific timeline positions.

When the `frame` parameter is omitted, automatic distribution rules apply:

- **1 image**: Used as the first frame.
- **2 images**: First and last frames.

**Examples**:

**Shorthand format:** When you don't need to specify a frame position, you can pass a plain image input directly.

```json
"frameImages": [
  "aac49721-1964-481a-ae78-8a4e29b91402"
]
```

**Object format:** When you need to specify a position, use an object with `image` and either `frame` or `timestamp` (model-dependent).

```json
"frameImages": [
  {
    "image": "aac49721-1964-481a-ae78-8a4e29b91402",
    "frame": "first"
  }
]
```

**First and last frames:** With two images, they automatically become the first and last frames of the video sequence. You can mix shorthand and object formats.

```json
"frameImages": [
  "aac49721-1964-481a-ae78-8a4e29b91402",
  {
    "image": "3ad204c3-a9de-4963-8a1a-c3911e3afafe",
    "frame": "last"
  }
]
```

**Format 1: string[]**:

- **Type**: `string`

Image input (UUID, URL, Data URI, or Base64).

**Format 2: object[]**:

#### [image](https://runware.ai/docs/models/minimax-h3#request-inputs-frameimages-format-2-image)

- **Path**: `inputs.frameImages.image`
- **Type**: `string`
- **Required**: true

Image input (UUID, URL, Data URI, or Base64).

#### [frame](https://runware.ai/docs/models/minimax-h3#request-inputs-frameimages-format-2-frame)

- **Path**: `inputs.frameImages.frame`
- **Type**: `object`

Target frame position for the image. Supports first and last frame.

**Allowed values**:

- `first` First frame of the video.
- `last` Last frame of the video.
- `0` Frame index 0 (first frame).
- `-1` Frame index -1 (last frame).

### [referenceVideos](https://runware.ai/docs/models/minimax-h3#request-inputs-referencevideos)

- **Path**: `inputs.referenceVideos`
- **Type**: `array of strings`
- **Min items**: `1`
- **Max items**: `3`
- **Size (each)**: `256 px to 5760 px`
- **Aspect ratio (each)**: `1:2.5 to 2.5:1`
- **Frame rate (each)**: `23.98 fps to 60 fps`
- **Duration (each)**: `2s to 15s`
- **File size (each)**: `up to 50 MB`

List of reference videos (UUID or URL).

### [referenceAudios](https://runware.ai/docs/models/minimax-h3#request-inputs-referenceaudios)

- **Path**: `inputs.referenceAudios`
- **Type**: `array of strings`
- **Min items**: `1`
- **Max items**: `3`
- **Duration (each)**: `2s to 15s`
- **File size (each)**: `up to 15 MB`
- **Duration (each)**: `2s to 15s`

List of reference audios (UUID or URL).

**Core Parameters**

Primary parameters that define the task output.

### [model](https://runware.ai/docs/models/minimax-h3#request-model)

- **Type**: `string`
- **Required**: true
- **Value**: `minimax:h3@0`

Send `auto` instead of an AIR identifier to have the model router choose the model for this task. The rest of the request is unchanged, every value you send is kept, and only the parameters you leave out are filled in. See [Model Router](https://runware.ai/docs/models-api/model-router).

**Learn more** (3 resources):

- [Text To Image: Model Selection](https://runware.ai/docs/learn/text-to-image#model-selection) (learn)
- [Image Inpainting: Model Specialized Inpainting Models](https://runware.ai/docs/learn/image-inpainting#model-specialized-inpainting-models) (learn)
- [Image Outpainting: Other Critical Parameters](https://runware.ai/docs/learn/image-outpainting#other-critical-parameters) (learn)

### [positivePrompt](https://runware.ai/docs/models/minimax-h3#request-positiveprompt)

- **Type**: `string`
- **Required**: true
- **Min**: `2`
- **Max**: `7000`

Text prompt describing elements to include in the generated output.

**Learn more** (1 resource):

- [Prompts](https://runware.ai/docs/learn/prompts) (learn)

### [width](https://runware.ai/docs/models/minimax-h3#request-width)

- **Type**: `integer`
- **Default**: `2560`
- **Paired with**: height

Width of the generated media in pixels.

**Learn more** (2 resources):

- [Dimensions](https://runware.ai/docs/learn/dimensions) (learn)
- [Image Outpainting: Dimensions Critical For Outpainting](https://runware.ai/docs/learn/image-outpainting#dimensions-critical-for-outpainting) (learn)

### [height](https://runware.ai/docs/models/minimax-h3#request-height)

- **Type**: `integer`
- **Default**: `1440`
- **Paired with**: width

Height of the generated media in pixels.

**Learn more** (2 resources):

- [Dimensions](https://runware.ai/docs/learn/dimensions) (learn)
- [Image Outpainting: Dimensions Critical For Outpainting](https://runware.ai/docs/learn/image-outpainting#dimensions-critical-for-outpainting) (learn)

### [resolution](https://runware.ai/docs/models/minimax-h3#request-resolution)

- **Type**: `string`

Resolution preset for the output. When used with input media, automatically matches the aspect ratio from the input.

**Allowed values**: `768p` `1440p`

### [duration](https://runware.ai/docs/models/minimax-h3#request-duration)

- **Type**: `integer`
- **Min**: `4`
- **Max**: `15`
- **Default**: `5`

Length of the generated video in seconds. The total number of frames produced is determined by duration multiplied by the model's frame rate (fps).

### [seed](https://runware.ai/docs/models/minimax-h3#request-seed)

- **Type**: `integer`
- **Min**: `0`
- **Max**: `4294967295`

Random seed for reproducible generation. When not provided, a random seed is generated in the unsigned 32-bit range.

**Settings**

Technical parameters to fine-tune the inference process. These must be nested inside the \`settings\` object.

### [promptExpansion](https://runware.ai/docs/models/minimax-h3#request-settings-promptexpansion)

- **Path**: `settings.promptExpansion`
- **Type**: `string`
- **Default**: `balanced`

Level of automatic prompt rewriting applied before generation. Affects reproducibility.

**Allowed values**:

- `disabled` No prompt rewriting.
- `balanced` Balanced rewriting.
- `quality` Quality-first rewriting, at the cost of longer generation time.

## Response Parameters

### [taskType](https://runware.ai/docs/models/minimax-h3#response-tasktype)

- **Type**: `string`
- **Required**: true
- **Value**: `videoInference`

Identifier for the type of task this response belongs to.

### [taskUUID](https://runware.ai/docs/models/minimax-h3#response-taskuuid)

- **Type**: `string`
- **Required**: true
- **Format**: `UUID v4`

UUID v4 identifier echoed from the original request, used to match async responses to their tasks.

### [videoUUID](https://runware.ai/docs/models/minimax-h3#response-videouuid)

- **Type**: `string`
- **Required**: true
- **Format**: `UUID v4`

UUID of the output video.

### [videoURL](https://runware.ai/docs/models/minimax-h3#response-videourl)

- **Type**: `string`
- **Format**: `uri`

URL of the output video.

### [videoBase64Data](https://runware.ai/docs/models/minimax-h3#response-videobase64data)

- **Type**: `string`

Base64-encoded video data.

### [videoDataURI](https://runware.ai/docs/models/minimax-h3#response-videodatauri)

- **Type**: `string`
- **Format**: `uri`

Data URI of the output video.

### [seed](https://runware.ai/docs/models/minimax-h3#response-seed)

- **Type**: `integer`

The seed used for generation. If none was provided, shows the randomly generated seed.

### [NSFWContent](https://runware.ai/docs/models/minimax-h3#response-nsfwcontent)

- **Type**: `boolean`

Flag indicating if NSFW content was detected.

### [cost](https://runware.ai/docs/models/minimax-h3#response-cost)

- **Type**: `float`

Task cost in USD. Present when `includeCost` is set to `true` in the request.

## Examples

### text-to-video (text-to-video)

[Watch video](https://assets.runware.ai/examples/minimax-h3/7bbe9f48-7002-4522-87c4-f9eae54c30d7.mp4)

```json
{
  "taskType": "videoInference",
  "taskUUID": "6592f33b-6e7e-4bcb-8973-836bfc3c92ff",
  "model": "minimax:h3@0",
  "positivePrompt": "Create a polished six-second living team portrait for a mountain ski patrol recruitment campaign. Single continuous landscape shot outside an alpine rescue hut at blue hour after fresh snowfall. Eight adult ski patrollers stand shoulder-to-shoulder facing the camera in practical red rescue jackets, dark snow pants, helmets, radios and avalanche gear; varied ages and appearances, relaxed but professional, no visible logos. Skis and a rescue toboggan rest neatly beside the timber equipment bay. Warm amber light spills from the hut against cool cobalt snow and a pale mountain sky.\n\nTemporal progression: begin with the complete patrol lineup holding a formal group-photo pose while one deliberate gap remains near the center. After about one second, a lively black-and-white avalanche rescue dog runs in from the right foreground, kicking up soft powder. The patrollers notice it, smile naturally and shift just enough to open the gap. The dog reaches the center, turns toward camera and sits proudly between two patrollers. One patroller briefly lowers a gloved hand to steady the dog. By the final two seconds, everyone is composed and looking into the lens as a cohesive team portrait, with the dog's breath faintly visible in the cold air and its tail making one final wag.\n\nCamera: eye-level medium-wide group framing, symmetrical but natural composition, all eight people visible head to boot, dog fully visible, mountain ridge retained above them. Locked tripod with an almost imperceptible slow push-in; no cuts, no reframing, no handheld shake. Preserve consistent faces, clothing, body count and equipment throughout. Premium documentary realism with subtle 1970s alpine tourism color influence, fine film grain, crisp winter air, restrained rescue-red, amber and deep-blue palette. Natural synchronized sound only: soft boots shifting in snow, the dog's paws and collar, a light mountain breeze and one quiet shared chuckle. No dialogue, captions, titles, signage, watermarks or added graphics.",
  "width": 1344,
  "height": 768,
  "duration": 6
}
```

---

### first-frame (first-frame)

[Watch video](https://assets.runware.ai/examples/minimax-h3/b7466efe-7e41-4d8e-8b74-c291bd4d9791.mp4)

```json
{
  "taskType": "videoInference",
  "taskUUID": "3cd47883-fa01-412f-8263-b4c3725aea6e",
  "model": "minimax:h3@0",
  "positivePrompt": "The scene continues naturally from the supplied first frame as a single uninterrupted eight-second stock-footage shot inside an indoor wasabi farm. The worker slowly loosens the wet gravel around the plant with one hand, then lifts the mature wasabi root smoothly into view, revealing its knobbly pale-green rhizome and fine dangling roots. Small pebbles tumble back into the shallow bed, droplets fall from the root, and nearby heart-shaped leaves sway gently from the movement and circulating greenhouse air. The camera performs a restrained, steady slider move forward and slightly to the right while maintaining the root and gloved hands as the focal point; realistic parallax develops across the receding crop rows. Finish with the freshly harvested rhizome held clearly in the foreground for useful editorial coverage. Photorealistic premium agricultural documentary b-roll, controlled natural motion, authentic horticultural handling, shallow depth of field, soft diffused skylight, cool humid atmosphere, rich emerald greens and muted silver-gray infrastructure, crisp water droplets, subtle filmic contrast. No cuts, no time lapse, no talking, no eye contact, no signage, no logos, no captions, no on-screen text. Natural synchronized ambience only: low irrigation trickle, faint ventilation hum, soft leaf rustle, gravel shifting, and droplets landing.",
  "resolution": "768p",
  "duration": 8,
  "inputs": {
    "frameImages": [
      {
        "image": "https://assets.runware.ai/assets/inputs/93bc5eb1-8f3d-4c56-9232-06f3da7dce28.jpg",
        "frame": "first"
      }
    ]
  }
}
```

---

### first-last-frame (first-last-frame)

[Watch video](https://assets.runware.ai/examples/minimax-h3/791ce3a7-6145-424c-b103-4892addcb2fb.mp4)

```json
{
  "taskType": "videoInference",
  "taskUUID": "d9a4a833-4f42-4561-b503-236bb89f2c02",
  "model": "minimax:h3@0",
  "positivePrompt": "Create a polished single-shot colorway reveal for a professional diver propulsion vehicle, transitioning coherently from the supplied first frame to the supplied last frame. Begin with the rescue-orange scooter perfectly still on its wet black-basalt plinth. Hold the opening composition briefly, then send a narrow band of refracted cyan light slowly across the product from left to right. As the light passes, only the orange outer shell changes progressively into deep kelp green, with a convincing premium factory finish and tiny beads of water catching the light. Preserve the scooter's exact silhouette, dimensions, seams, fasteners, control module, black handles, ducted propeller, orientation, and position throughout; this is a product recolor, not a redesign or transformation. Keep the camera locked at the same eye-level three-quarter angle with no zoom, cuts, or reframing. Add restrained motion only in the environment: drifting microbubbles, gently moving aquatic caustic reflections, and a faint curl of cool mist along the plinth. Finish cleanly on the supplied kelp-green endpoint image and hold long enough to inspect the completed colorway. Photorealistic high-end industrial product cinematography, crisp materials, dark cyan expedition-lab atmosphere, controlled cool rim light, realistic reflections, no people, no logos, no lettering, no packaging, no extra objects. Synchronized sound: subtle underwater pressure ambience, delicate bubbles, and one soft low-frequency sweep following the moving light band; no voice and no music.",
  "resolution": "1440p",
  "duration": 10,
  "inputs": {
    "frameImages": [
      {
        "image": "https://assets.runware.ai/assets/inputs/5910fd92-f212-4f8f-b55b-d07d91486788.jpg",
        "frame": "first"
      },
      {
        "image": "https://assets.runware.ai/assets/inputs/7aa2c02b-988f-4012-882a-dc7172d2e571.jpg",
        "frame": "last"
      }
    ]
  }
}
```

---

### reference-to-video (reference-to-video)

[Watch video](https://assets.runware.ai/examples/minimax-h3/95422500-b456-43fe-8d7c-bec2011e2f58.mp4)

```json
{
  "taskType": "videoInference",
  "taskUUID": "50c2c534-6240-4f75-a203-70dcbddc2b5b",
  "model": "minimax:h3@0",
  "positivePrompt": "Create a polished 12-second municipal public-health talking-head video in a cinematic documentary style. The presenter is the wastewater scientist shown in the reference image: preserve her facial identity, short natural curls, warm brown skin, clear safety glasses, navy utility coverall, teal gloves, and approachable professional presence. Place her at a clean stainless-steel testing bench in the same modern wastewater laboratory, with softly defocused sample bottles and turquoise-lit utility pipes behind her. Use the reference image as the primary guide for subject identity, wardrobe, laboratory design, cyan-and-amber palette, and lighting.\n\nUse the reference video as guidance for framing and performance: one uninterrupted eye-level medium shot, presenter centered slightly left with modest negative space, restrained documentary camera language, and a very slow subtle push-in. Match its natural posture, direct eye contact, realistic blinking, small open-hand emphasis gesture, and reassuring finishing nod. Avoid exaggerated gesturing or commercial-presenter energy.\n\nUse the reference audio as the exact spoken performance and synchronize the presenter’s mouth, facial articulation, pauses, breath, gesture timing, and final nod precisely to it. Preserve a credible adult female voice and clean speech intelligibility. Add only subtle, realistic laboratory room tone underneath—soft ventilation and distant water-system hum—with no music and no competing machinery noise.\n\nThe mood is practical, trustworthy, and quietly urgent rather than alarming. Keep skin texture, hands, safety glasses, lip motion, and eye movement photorealistic. Maintain continuity throughout with no cuts, no additional people, no logos, no captions, no lower thirds, no visible brand names, and no on-screen text.",
  "resolution": "768p",
  "duration": 12,
  "inputs": {
    "referenceImages": [
      "https://assets.runware.ai/assets/inputs/40514195-8f31-4c96-90c2-745681242642.jpg"
    ],
    "referenceVideos": [
      "https://assets.runware.ai/assets/inputs/2979c59c-60b2-4eb5-b38b-ef5e2f115ed3.mp4"
    ],
    "referenceAudios": [
      "https://assets.runware.ai/assets/inputs/4723fc05-9966-41c2-b7e8-40ef4cc59920.mp3"
    ]
  }
}
```