Common Conventions

All interfaces require request headers Authorization: Bearer <apiKey> and Topview-Uid: <userId>

Successful response unified structure: { "code": "200", "message": "Success", "result": { ... } }

After task submission, returns taskId. Frontend polls the query interface to get results

I. Image-to-Video V2

1.1 Submit Task

POST /v2/common_task/image2video/task/submit

Request Headers

Parameter	Type	Required	Description
`Topview-Uid`	string	Yes	User ID

Request Body (JSON)

Parameter	Type	Required	Description
`model`	string	No	Model display name, e.g., `"Seedance 1.5 pro"`, see supported models table below
`firstFrameFileId`	string	No	File ID of the first frame image
`endFrameFileId`	string	No	File ID of the last frame image
`referenceImageFileIds`	string[]	No	List of reference image IDs in multi-image mode (at least 2, excluding first and last frames)
`prompt`	string	No	Prompt text
`aspectRatio`	string	No	Aspect ratio, e.g., `"16:9"`, `"9:16"`
`resolution`	integer	No	Resolution, `720` or `1080`
`duration`	integer	No	Video duration in seconds
`sound`	string	No	Whether to include sound effects, `"on"` or `"off"`, default `"off"`
`generatingCount`	integer	No	Number of videos to generate, default 1
`boardId`	string	No	Board ID. When provided, a task record will be created in the Board

Supported Models and Parameter Constraints

Note: The model parameter should use the model display name (e.g., "Seedance 1.5 pro"), not the code.

Model Name	Image Input Mode	Aspect Ratio Options (aspectRatio)	Resolution (resolution)	Duration (duration)	Native Audio	Remarks
Standard(Seedance 2.0)	First & Last Frame	Determined by image	480, 720,1080	4–15 seconds	✅	First and last frames must have same aspect ratio; supports internet search
Fast(Seedance 2.0 Fast)	First & Last Frame	Determined by image	480, 720	4–15 seconds	✅	First and last frames must have same aspect ratio; supports internet search
Seedance 1.5 pro	First & Last Frame	Determined by image	720, 1080	4–12 seconds	✅
Seedance 1.0 Pro Fast	Single Image	Determined by image	720, 1080	5, 10, 12 seconds	❌
Seedance 1.0 Pro	First & Last Frame	Determined by image	720, 1080	5, 10, 12 seconds	❌
Kling V3	First & Last Frame	Determined by image	720, 1080	3–15 seconds	✅	Supports multi-shot (up to 6 shots, total ≤15s); image short side ≥300px
Kling V3 Reference Video	Multi-Image (Reference)	9:16, 1:1, 16:9	720, 1080	3–15 seconds	✅	Supports multi-shot; image short side ≥300px
Kling O3	First & Last Frame	Determined by image	720, 1080	3–15 seconds	✅	Supports multi-shot (up to 6 shots, total ≤15s); image short side ≥300px
Kling O3 Reference-to-Video	Multi-Image (Reference, max 7)	9:16, 1:1, 16:9	720, 1080	3–15 seconds	✅	Supports @mention reference images; supports multi-shot
Kling 2.6	Single Image	Determined by image	N/A	5, 10 seconds	✅
Kling O1 Reference-to-Video	Multi-Image (Reference, max 7)	9:16, 1:1, 16:9	1080	3–10 seconds	❌	Supports @mention reference images; image parameters optional
Kling 2.5 Turbo Pro	First & Last Frame	Determined by image	1080	5, 10 seconds	❌
Kling 2.5 Turbo Std	Single Image	Determined by image	1080	5, 10 seconds	❌
Sora 2	Single Image	9:16, 16:9	N/A	4, 8, 12 seconds	❌	Does not support generating realistic humans
Sora 2 Pro	Single Image	9:16, 16:9	720, 1080	4, 8, 12 seconds	❌	Does not support generating realistic humans
Veo 3.1	First & Last Frame	9:16, 16:9	720, 1080, 2160	4, 6, 8 seconds (2160p only 8s)	✅
Veo 3.1 Reference to video	Multi-Image (Reference, max 3)	9:16, 16:9	720, 1080, 2160	8 seconds	✅
Veo 3.1 Fast	First & Last Frame	9:16, 16:9	720, 1080, 2160	4, 6, 8 seconds (2160p only 8s)	✅
Veo 3.1 Fast Reference to video	Multi-Image (Reference, max 3)	9:16, 16:9	720, 1080, 2160	8 seconds	✅
MiniMax-Hailuo-02	First & Last Frame	Determined by image	768, 1080	6, 10 seconds (1080p only 6s)	❌
MiniMax-Hailuo-2.3	Single Image	Determined by image	768, 1080	6, 10 seconds (1080p only 6s)	❌
MiniMax-Hailuo-2.3-Fast	Single Image	Determined by image	768, 1080	6, 10 seconds (1080p only 6s)	❌
Vidu Q3 Pro	Single Image	9:16, 3:4, 1:1, 4:3, 16:9	540, 720, 1080	1–16 seconds	✅
Vidu Q2 Reference to Video	Multi-Image (Reference, max 7)	9:16, 3:4, 1:1, 4:3, 16:9	1080	2–10 seconds	❌
Wan 2.6	Single Image	Determined by image	720, 1080	5, 10, 15 seconds	✅
Grok Video	Single Image	9:16, 3:4, 2:3, 1:1, 4:3, 3:2, 16:9	480, 720	1–15 seconds	❌
Topview Pro	Single Image	9:16, 3:4, 1:1, 4:3, 16:9	N/A	5, 10 seconds	❌
Topview Plus	Single Image	9:16, 1:1, 16:9	N/A	5, 10 seconds	❌
Topview Best	Single Image	9:16, 1:1, 16:9	N/A	5, 10 seconds	❌

Image input mode description:
- Single Image: Only firstFrameFileId takes effect
- First & Last Frame: Can pass both firstFrameFileId (first frame) and endFrameFileId (last frame)
- Multi-Image (Reference): Pass reference image list via referenceImageFileIds
sound: "on" or "off", default "off"; models supporting Native Audio can generate native audio when enabled
generatingCount: 1–4, default 1

Request Example

POST /v2/common_task/image2video/task/submit
Topview-Uid: user_123

{
  "model": "Seedance 1.5 pro",
  "firstFrameFileId": "file_abc123",
  "prompt": "A product slowly rotating on a clean white background, cinematic lighting",
  "aspectRatio": "16:9",
  "resolution": 1080,
  "duration": 5,
  "sound": "off",
  "generatingCount": 2
}

Response Example

{
  "code": "200",
  "message": "Success",
  "result": {
    "taskId": "task_i2v_v2_20250529001",
    "status": "init",
    "boardTaskIds": null,
    "boardId": null,
    "noticeUuid": "uuid_abc"
  }
}

1.2 Query Task Result

GET /v2/common_task/image2video/task/query

Request Parameters

Parameter	Type	Required	Description
`taskId`	string	Yes	Task ID returned from submission
`needCloudFrontUrl`	boolean	No	Whether to return CloudFront URL, default `false`

Request Example

GET /v2/common_task/image2video/task/query?taskId=task_i2v_v2_20250529001
Topview-Uid: user_123

Response Example (In Progress)

{
  "code": "200",
  "message": "Success",
  "result": {
    "taskId": "task_i2v_v2_20250529001",
    "status": "running",
    "errorMsg": null,
    "costCredit": "0.00",
    "videos": []
  }
}

Response Example (Success)

{
  "code": "200",
  "message": "Success",
  "result": {
    "taskId": "task_i2v_v2_20250529001",
    "status": "success",
    "errorMsg": null,
    "costCredit": "5.00",
    "videos": [
      {
        "status": "success",
        "errorMsg": null,
        "fileId": "file_video_001",
        "filePath": "https://cdn.example.com/videos/output_001.mp4",
        "width": 1920,
        "height": 1080,
        "duration": 5000,
        "coverPath": "https://cdn.example.com/covers/cover_001.jpg",
        "boardTaskId": null
      },
      {
        "status": "success",
        "errorMsg": null,
        "fileId": "file_video_002",
        "filePath": "https://cdn.example.com/videos/output_002.mp4",
        "width": 1920,
        "height": 1080,
        "duration": 5000,
        "coverPath": "https://cdn.example.com/covers/cover_002.jpg",
        "boardTaskId": null
      }
    ]
  }
}

II. Text-to-Video

2.1 Submit Task

POST /v1/common_task/text2video/task/submit

Request Headers

Parameter	Type	Required	Description
`Topview-Uid`	string	Yes	User ID

Request Body (JSON)

Parameter	Type	Required	Description
`model`	string	Yes	Model name, e.g., `"Seedance 1.5 pro"`, `"Kling V3"`, `"Sora 2"`, `"Topview Pro"`
`prompt`	string	Yes	Positive prompt text
`aspectRatio`	string	No	Aspect ratio, e.g., `"16:9"`, `"9:16"`, `"1:1"`
`resolution`	integer	No	Resolution, `720` or `1080`
`duration`	integer	No	Video duration in seconds
`sound`	string	No	Whether to include sound effects, `"on"` or `"off"`
`generatingCount`	integer	No	Number of generations, `1~4`
`boardId`	string	No	Board ID. When provided, a task record will be created in the Board

Supported Models and Parameter Constraints

Note: The model parameter should use the model display name (e.g., "Seedance 1.5 pro"), not the code.

Model Name	Aspect Ratio Options (aspectRatio)	Resolution (resolution)	Duration (duration)	Native Audio	Remarks
Standard(Seedance 2.0)	First & Last Frame	Determined by image	480, 720, 1080	4–15 seconds	✅	First and last frames must have same aspect ratio; supports internet search
Fast(Seedance 2.0 Fast)	First & Last Frame	Determined by image	480, 720	4–15 seconds	✅	First and last frames must have same aspect ratio; supports internet search
Seedance 1.5 pro	9:16, 3:4, 1:1, 4:3, 16:9	720, 1080	4–12 seconds	✅
Seedance 1.0 Pro Fast	9:16, 3:4, 1:1, 4:3, 16:9	720, 1080	5, 10, 12 seconds	❌
Seedance 1.0 Pro	9:16, 3:4, 1:1, 4:3, 16:9	720, 1080	5, 10, 12 seconds	❌
Kling O3	9:16, 1:1, 16:9	720, 1080	3–15 seconds	✅	Supports multi-shot (up to 6 shots, total ≤15s)
Kling V3	9:16, 1:1, 16:9	720, 1080	3–15 seconds	✅	Supports multi-shot (up to 6 shots, total ≤15s)
Kling 2.6	9:16, 1:1, 16:9	N/A	5, 10 seconds	✅
Kling 2.5 Turbo Pro	9:16, 1:1, 16:9	1080	5, 10 seconds	❌
Kling 2.5 Turbo Std	9:16, 1:1, 16:9	1080	5, 10 seconds	❌
Sora 2	9:16, 16:9	N/A	4, 8, 12 seconds	❌
Sora 2 Pro	9:16, 16:9	720, 1080	4, 8, 12 seconds	❌
Veo 3.1	9:16, 16:9	720, 1080, 2160	4, 6, 8 seconds (2160p only 8s)	✅
Veo 3.1 Fast	9:16, 16:9	720, 1080, 2160	4, 6, 8 seconds (2160p only 8s)	✅
MiniMax-Hailuo-02	Determined by model	768, 1080	6, 10 seconds (1080p only 6s)	❌
MiniMax-Hailuo-2.3	Determined by model	768, 1080	6, 10 seconds (1080p only 6s)	❌
Vidu Q3 Pro	9:16, 3:4, 1:1, 4:3, 16:9	540, 720, 1080	1–16 seconds	✅
Vidu Q2	9:16, 3:4, 1:1, 4:3, 16:9	1080	2–10 seconds	❌
Wan 2.6	9:16, 3:4, 1:1, 4:3, 16:9	720, 1080	5, 10, 15 seconds	✅
Grok Video	9:16, 3:4, 2:3, 1:1, 4:3, 3:2, 16:9	480, 720	1–15 seconds	❌
Topview Pro	9:16, 3:4, 1:1, 4:3, 16:9	N/A	5, 10 seconds	❌
Topview Plus	9:16, 1:1, 16:9	N/A	5, 10 seconds	❌
Topview Best	9:16, 1:1, 16:9	N/A	5, 10 seconds	❌

sound: "on" or "off", default "off"; models supporting Native Audio can generate native audio when enabled
generatingCount: 1–4 (required)

Request Example

POST /v1/common_task/text2video/task/submit
Topview-Uid: user_123

{
  "model": "Seedance 1.5 pro",
  "prompt": "A futuristic city at night with neon lights reflecting on wet streets, cinematic 4K",
  "aspectRatio": "16:9",
  "resolution": 1080,
  "duration": 5,
  "sound": "on",
  "generatingCount": 1
}

Response Example

{
  "code": "200",
  "message": "Success",
  "result": {
    "taskId": "task_t2v_20251128001",
    "status": "init",
    "noticeUuid": "uuid_xyz"
  }
}

2.2 Query Task Result

GET /v1/common_task/text2video/task/query

Request Parameters

Parameter	Type	Required	Description
`taskId`	string	Yes	Task ID returned from submission
`needCloudFrontUrl`	boolean	No	Whether to return CloudFront URL, default `false`

Request Example

GET /v1/common_task/text2video/task/query?taskId=task_t2v_20251128001
Topview-Uid: user_123

Response Example (Success)

{
  "code": "200",
  "message": "Success",
  "result": {
    "taskId": "task_t2v_20251128001",
    "status": "success",
    "errorMsg": null,
    "costCredit": "10.00",
    "videos": [
      {
        "status": "success",
        "errorMsg": null,
        "fileId": "file_video_t2v_001",
        "filePath": "https://cdn.example.com/videos/t2v_output_001.mp4",
        "width": 1920,
        "height": 1080,
        "duration": 5000,
        "coverPath": "https://cdn.example.com/covers/t2v_cover_001.jpg"
      }
    ]
  }
}

Response Example (Failure)

{
  "code": "200",
  "message": "Success",
  "result": {
    "taskId": "task_t2v_20251128001",
    "status": "fail",
    "errorMsg": "Model service unavailable",
    "costCredit": "0.00",
    "videos": []
  }
}

III. Omni Reference

Omni Reference supports generating new videos through reference images/videos + prompt. The prompt can reference uploaded materials using the <<<ImageN>>> / <<<VideoN>>> syntax.

3.1 Submit Task

POST /v1/common_task/omni_reference/task/submit

Request Headers

Parameter	Type	Required	Description
`Topview-Uid`	string	Yes	User ID

Request Body (JSON)

Parameter	Type	Required	Description
`model`	string	Yes	Model name, e.g., `"Standard"`, `"Fast"`, `"Kling O3 Video-Edit"`, `"Kling O1 Video-Edit"`
`prompt`	string	Yes	Prompt text, supports `<<<Image1>>>`, `<<<Video1>>>` and other reference syntax
`inputImages`	object[]	No	List of reference images, see table below
`inputVideos`	object[]	No	List of reference videos, see table below
`aspectRatio`	string	No	Aspect ratio, e.g., `"16:9"`
`resolution`	integer	No	Resolution, `480`, `720` or `1080`
`duration`	integer	No	Video duration in seconds
`generatingCount`	integer	No	Number of generations, `1~4`
`internetSearch`	boolean	No	Whether to enable internet search, default `false`
`boardId`	string	No	Board ID. When provided, a task record will be created in the Board

inputImages Object Fields

Parameter	Type	Required	Description
`fileId`	string	No	File ID returned from upload interface
`inputImageS3Path`	string	No	Image S3 path (choose one with fileId)
`name`	string	Yes	Reference name, e.g., `"Image1"`, corresponds to `<<<Image1>>>` in prompt

inputVideos Object Fields

Parameter	Type	Required	Description
`fileId`	string	No	File ID returned from upload interface
`inputVideoS3Path`	string	No	Video S3 path (choose one with fileId)
`name`	string	Yes	Reference name, e.g., `"Video1"`, corresponds to `<<<Video1>>>` in prompt

Supported Models and Parameter Constraints

Note: The model parameter should use the model display name (e.g., "Standard", "Kling O3 Video-Edit"), not the code.

(Video Edit Mode)

Model Name	Aspect Ratio Options (aspectRatio)	Resolution (resolution)	Duration (duration)	Native Audio	Internet Search
Standard (Seedance 2.0)	9:16, 3:4, 1:1, 4:3, 16:9, 21:9	480, 720, 1080	4–15 seconds	✅	✅
Fast(Seedance 2.0 Fast)	9:16, 3:4, 1:1, 4:3, 16:9, 21:9	480, 720	4–15 seconds	✅	✅

Seedance 2.0 mixed upload limits (inputImages + inputVideos combined):

Maximum 9 images
Maximum 3 videos (single video total duration ≤15.4s)
Maximum 3 audios (single audio total duration ≤15.4s)
Images + videos + audios combined cannot exceed 12 items

Kling (Video Edit Mode)

Model Name	Edit Mode (videoEditMode)	Reference Images	Resolution (resolution)	Keep Original Audio	Remarks
Kling O3 Video-Edit	base, feature	Maximum 4 (optional)	720, 1080	✅	Supports @mention reference images/videos
Kling O1 Video-Edit	base, feature	Maximum 4 (optional)	1080	✅	Supports @mention reference images/videos; billed per second

Kling input video requirements (inputVideos, inputVideoMode: singleVideo):

Width/Height: 720–2160px
Video duration: 3–10 seconds

Grok Video (Video Edit Mode)

Model Name	Input Videos	aspectRatio / resolution / duration	Remarks
Grok Video	Exactly 1	Do not pass (determined by input)	`<<<VideoName>>>` reference prompt; no images

Grok input video requirements (inputVideos, inputVideoMode: singleVideo):

Width/Height: 300–2160px
Video duration: 1–15 seconds
Do not pass inputImages, internetSearch, aspectRatio, resolution, or duration

Request Example (Grok Video Edit)

POST /v1/common_task/omni_reference/task/submit
Topview-Uid: user_123

{
  "model": "Grok Video",
  "prompt": "Apply cinematic color grading to <<<Video1>>>",
  "inputVideos": [
    {
      "fileId": "file_video_001",
      "name": "Video1"
    }
  ],
  "generatingCount": 1
}

Common parameters description:

generatingCount: 1–4 (required)

Request Example (Image Reference)

POST /v1/common_task/omni_reference/task/submit
Topview-Uid: user_123

{
  "model": "Kling O3 Video-Edit",
  "prompt": "Transform <<<Image1>>> into a watercolor painting style animation with gentle movement",
  "inputImages": [
    {
      "fileId": "file_img_abc001",
      "name": "Image1"
    }
  ],
  "aspectRatio": "16:9",
  "resolution": 720,
  "duration": 5,
  "generatingCount": 1,
  "internetSearch": false
}

Request Example (Image + Video Mixed Reference)

POST /v1/common_task/omni_reference/task/submit
Topview-Uid: user_123

{
  "model": "Standard",
  "prompt": "Apply the style from <<<Image1>>> to the motion in <<<Video1>>>",
  "inputImages": [
    {
      "fileId": "file_img_style001",
      "name": "Image1"
    }
  ],
  "inputVideos": [
    {
      "fileId": "file_video_motion001",
      "name": "Video1"
    }
  ],
  "aspectRatio": "9:16",
  "resolution": 1080,
  "duration": 5,
  "generatingCount": 2
}

Response Example

{
  "code": "200",
  "message": "Success",
  "result": {
    "taskId": "task_omni_20260302001",
    "status": "init",
    "noticeUuid": "uuid_omni_abc"
  }
}

3.2 Query Task Result

GET /v1/common_task/omni_reference/task/query

Request Parameters

Parameter	Type	Required	Description
`taskId`	string	Yes	Task ID returned from submission
`needCloudFrontUrl`	boolean	No	Whether to return CloudFront URL, default `false`

Request Example

GET /v1/common_task/omni_reference/task/query?taskId=task_omni_20260302001
Topview-Uid: user_123

Response Example (Success)

{
  "code": "200",
  "message": "Success",
  "result": {
    "taskId": "task_omni_20260302001",
    "status": "success",
    "errorMsg": null,
    "costCredit": "8.00",
    "videos": [
      {
        "status": "success",
        "errorMsg": null,
        "fileId": "file_video_omni_001",
        "filePath": "https://cdn.example.com/videos/omni_output_001.mp4",
        "width": 1920,
        "height": 1080,
        "duration": 5000,
        "coverPath": "https://cdn.example.com/covers/omni_cover_001.jpg",
        "boardTaskId": null
      }
    ]
  }
}

IV. Common Specifications

Task Status Enumeration

Status Value	Meaning
`init`	Task submitted, waiting for processing
`running`	Task is being processed
`success`	Task completed successfully
`fail`	Task failed, see `errorMsg` field for error reason

Polling Recommendations

Recommended polling interval: 3–5 seconds, timeout should be set to 5 minutes.

Sub-tasks in `videos`

When generatingCount > 1, the videos list contains multiple sub-task results. Each sub-task has an independent status. When the parent task status=success, all sub-tasks are completed (some sub-tasks may be fail, check the sub-task's status and errorMsg).

Omni Reference Prompt Reference Syntax

Reference input materials in prompt using <<<ImageN>>> or <<<VideoN>>>:

"prompt": "Apply the color style from <<<Image1>>> to the scene and follow the motion from <<<Video1>>>"

Where Image1 corresponds to the item with name="Image1" in inputImages, and Video1 is similar.

V. Credit Billing Explanation

totalCost = unitCost × generatingCount

unitCost Calculation Rules:

General Models: Query pricing table using resolution, duration, sound parameters
- Models with resolution parameter: key = {resolution}_{duration}[_sound], e.g., 1080_10_sound
- Models without resolution parameter: key = {duration}[_sound], e.g., 10_sound
Kling O1 Video-Edit (special): unitCost = 0.8 × ceil(input video seconds)
Grok Video-Edit (special): unitCost = 0.8 × ceil(input video seconds)
Grok Video (I2V / T2V): unitCost = 0.8 × duration (480p and 720p same rate)

5.1 Model Pricing Tables (Unit: Credits/Generation, generatingCount=1)

Standard(Seedance 2.0) / Fast(Seedance 2.0 Fast)

Billing parameters: resolution, duration

Enabling sound does not add cost. Credit value = per-second rate × duration (seconds).

Image-to-Video / Text-to-Video / Video-Edit (without video input)

Model	Resolution	Credits per second
Standard(Seedance 2.0)	1080p	2.5
Standard(Seedance 2.0)	720p	1
Standard(Seedance 2.0)	480p	0.5
Fast(Seedance 2.0 Fast)	720p	0.8
Fast(Seedance 2.0 Fast)	480p	0.4

Video-Edit (with video input)

Model	Resolution	Credits per second
Standard(Seedance 2.0)	1080p	3
Standard(Seedance 2.0)	720p	1.2
Standard(Seedance 2.0)	480p	0.6
Fast(Seedance 2.0 Fast)	720p	0.96
Fast(Seedance 2.0 Fast)	480p	0.48

Seedance 1.5 pro

Billing parameters: resolution, duration, sound Enabling sound approximately ×2.1

Resolution	4s	5s	6s	7s	8s	9s	10s	11s	12s
720p	0.48	0.60	0.72	0.84	0.96	1.08	1.20	1.32	1.44
720p + sound	1.00	1.25	1.50	1.75	2.00	2.25	2.50	2.75	3.00
1080p	1.12	1.40	1.68	1.96	2.24	2.52	2.80	3.08	3.36
1080p + sound	2.20	2.75	3.30	3.85	4.40	4.95	5.50	6.05	6.60

Seedance 1.0 Pro Fast / Seedance 1.0 Pro

Billing parameters: resolution, duration (does not support sound)

Model	Resolution	5s	10s	12s
1.0 Pro Fast	720p	0.35	0.70	0.84
1.0 Pro Fast	1080p	0.75	1.50	1.80
1.0 Pro	720p	1.00	2.00	2.40
1.0 Pro	1080p	2.25	4.50	5.40

Kling V3 / Kling V3 Reference Video

Billing parameters: resolution, duration, sound Enabling sound: 720p approximately +60%, 1080p approximately +25%

Resolution	3s	5s	8s	10s	12s	15s
720p	1.5	2.5	4.0	5.0	6.0	7.5
720p + sound	2.4	4.0	6.4	8.0	9.6	12.0
1080p	2.4	4.0	6.4	8.0	9.6	12.0
1080p + sound	3.0	5.0	8.0	10.0	12.0	15.0

Kling O3 / Kling O3 Reference-to-Video

Billing parameters: resolution, duration, sound Enabling sound approximately +12.5%

Resolution	3s	5s	8s	10s	12s	15s
720p	1.5	2.5	4.0	5.0	6.0	7.5
720p + sound	1.8	3.0	4.8	6.0	7.2	9.0
1080p	2.4	4.0	6.4	8.0	9.6	12.0
1080p + sound	2.7	4.5	7.2	9.0	10.8	13.5

Kling 2.6

Billing parameters: duration, sound (no resolution dimension) Enabling sound approximately ×2.1

Duration	No Sound	With Sound
5s	1.55	3.25
10s	3.10	6.50

Kling O1 Reference-to-Video

Billing parameters: duration (1080p only, does not support sound) Approximately 0.5 credits/second

3s	4s	5s	6s	7s	8s	9s	10s
1.5	2.0	2.5	3.0	3.5	4.0	4.5	5.0

Kling 2.5 Turbo Pro / Kling 2.5 Turbo Std

Billing parameters: duration (1080p only, does not support sound)

Model	5s	10s
Kling 2.5 Turbo Pro	1.55	3.10
Kling 2.5 Turbo Std	1.00	2.00

Sora 2 / Sora 2 Pro

Billing parameters: resolution, duration (does not support sound)

Model	Resolution	4s	8s	12s
Sora 2	N/A	2.24	4.48	6.72
Sora 2 Pro	720p	6.72	13.44	20.16
Sora 2 Pro	1080p	11.20	22.40	33.60

Veo 3.1 / Veo 3.1 Reference to video

Billing parameters: resolution, duration, sound 720p and 1080p same price; enabling sound ×2; 2160p only supports 8s

Resolution	4s	4s+sound	6s	6s+sound	8s	8s+sound
720p / 1080p	4.4	8.8	6.6	13.2	8.8	17.6
2160p	—	—	—	—	17.6	26.4

Veo 3.1 Fast / Veo 3.1 Fast Reference to video

Billing parameters: resolution, duration, sound 720p and 1080p same price; enabling sound approximately +50%; 2160p only supports 8s

Resolution	4s	4s+sound	6s	6s+sound	8s	8s+sound
720p / 1080p	2.4	3.6	3.6	5.4	4.8	7.2
2160p	—	—	—	—	13.6	16.0

MiniMax-Hailuo-02 / MiniMax-Hailuo-2.3

Billing parameters: resolution, duration (does not support sound) 1080p only supports 6s

Resolution	6s	10s
768p	1.6	3.2
1080p	2.8	—

MiniMax-Hailuo-2.3-Fast

Billing parameters: resolution, duration (does not support sound) 1080p only supports 6s

Resolution	6s	10s
768p	1.1	1.8
1080p	1.8	—

Vidu Q3 Pro

Billing parameters: resolution, duration, sound Enabling sound does not add cost; supports 1–16s any duration, linear billing

Resolution	Unit Price (Credits/sec)	5s	10s	16s
540p	0.40	2.0	4.0	6.4
720p	0.90	4.5	9.0	14.4
1080p	1.00	5.0	10.0	16.0

Vidu Q2 (T2V) / Vidu Q2 Reference to Video (I2V)

Billing parameters: duration (1080p only, does not support sound) I2V Reference mode includes approximately 2.1 credits base cost, same duration approximately ×2 of T2V

Model	2s	4s	6s	8s	10s
Vidu Q2 (T2V)	0.84	1.40	1.96	2.52	3.08
Vidu Q2 Reference to Video (I2V)	2.38	2.94	3.50	4.06	4.62

Wan 2.6 (I2V and T2V same price)

Billing parameters: resolution, duration, sound Enabling sound does not add cost

Resolution	5s	10s	15s
720p	2.3	4.6	6.9
1080p	3.85	7.7	11.55

Topview Series (I2V and T2V same price)

Billing parameters: duration (no resolution, does not support sound)

Model	5s	10s
Pro	1.0	2.0
Plus	2.7	5.4
Best	7.75	15.5

Kling O3 Video-Edit (Video Edit Mode)

Billing parameters: resolution, output video duration (determined by input video, max 10s, does not support sound)

Resolution	Unit Price (Credits/sec)	3s	5s	8s	10s
720p	1.70	5.1	8.5	13.6	17.0
1080p	2.00	6.0	10.0	16.0	20.0

Kling O1 Video-Edit (Video Edit Mode)

Billing parameters: Input video duration (rounded up per second, does not support sound)

unitCost = 0.8 × ceil(input video seconds)

3s	4s	5s	6s	7s	8s	9s	10s
2.4	3.2	4.0	4.8	5.6	6.4	7.2	8.0

Grok Video (Image-to-Video / Text-to-Video)

Billing parameters: resolution, duration (480p and 720p same per-second rate; no sound)

unitCost = 0.8 × duration (seconds)

Resolution	Credits per second
480p	0.8
720p	0.8

Example: 10s @ 720p → 8.0 credits per generation.

Grok Video (Video Edit Mode)

Billing parameters: Input video duration (rounded up per second)

unitCost = 0.8 × ceil(input video seconds)

1s	5s	10s	15s
0.8	4.0	8.0	12.0

5.2 Native Audio (sound) Surcharge Reference

Surcharge	Models
No additional cost	Standard/Fast, Vidu Q3 Pro, Wan 2.6
Approximately +12.5%	Kling O3, Kling O3 Reference-to-Video
Approximately +25% (1080p) / +60% (720p)	Kling V3, Kling V3 Reference Video
Approximately +50%	Veo 3.1 Fast series, Veo 3.1 2160p
×2	Seedance 1.5 pro, Kling 2.6, Veo 3.1 (720p/1080p)
Not supported	Seedance 1.0 Pro series, Kling O1/2.5 series, Sora 2 series, MiniMax series, Vidu Q2, Topview full series, Grok Video