Wan 2.6 Text-To-Video Spark Generation

Generate videos from text prompts using the wan2.6-t2v-spark model.

POST /vendors/mulerouter/v1/wan2.6-t2v-spark/generation

Create Generation Task

cURL

curl --request POST \
  --url https://api.mulerun.com/vendors/mulerouter/v1/wan2.6-t2v-spark/generation \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "prompt": "<string>",
  "negative_prompt": "<string>",
  "size": "1920*1080",
  "prompt_extend": true,
  "shot_type": "single",
  "audio": true,
  "audio_url": "<string>",
  "seed": 1073741823
}
'

202

{
  "task_info": {
    "id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
    "created_at": "2023-11-07T05:31:56Z",
    "updated_at": "2023-11-07T05:31:56Z"
  }
}

This is a fine-tuned model by MuleRouter based on the Wan2.6 Text-to-Video architecture. It offers improved quality and performance for specific use cases.

Overview

Generate videos from text prompts using the wan2.6-t2v-spark model with support for longer durations and multi-shot generation.

Key Features

• Text-to-video generation with audio support
• 1080P resolution options
• 5s, 10s, or 15s duration
• Single or multi-shot generation
• Fine-tuned for enhanced visual quality

Resolution Options

1080P

• 1920×1080 (16:9)
• 1080×1920 (9:16)
• 1440×1440 (1:1)
• 1632×1248 (4:3)
• 1248×1632 (3:4)

Example Requests

Basic Text-to-Video

{
  "prompt": "Miyazaki-style mule dancing",
  "size": "1920*1080",
  "duration": 5
}

Multi-shot Video

{
  "prompt": "Waves crashing against rocks, water splashing, sunlight on the sea",
  "size": "1920*1080",
  "duration": 10,
  "prompt_extend": true,
  "multi_shots": "multi",
  "audio": true
}

With Custom Audio

{
  "prompt": "A person walking through a city at night",
  "size": "1920*1080",
  "duration": 15,
  "audio_url": "https://example.com/city_sounds.mp3"
}

Parameters

multi_shots

• Default: single
• Effect: When prompt rewriting is enabled, controls whether the output is single-shot or multi-shot
• Only effective when: prompt_extend is enabled

duration

• Options: 5, 10, or 15 seconds
• Note: Longer durations may increase processing time

Authorizations

Authorization · string · header · required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Body

application/json

prompt · string · required

Text description for the desired video content (max 2000 characters).

Maximum string length: 2000

negative_prompt · string

Negative prompt describing unwanted content (max 500 characters).

Maximum string length: 500

size · enum<string> · default: 1920*1080

Output resolution ("width*height"). Supported tiers:

• 1080P: 1920*1080 (16:9), 1080*1920 (9:16), 1440*1440 (1:1), 1632*1248 (4:3), 1248*1632 (3:4)

Available options: 1920*1080, 1080*1920, 1440*1440, 1632*1248, 1248*1632

duration · enum<integer>

Video duration in seconds. Supported values 5, 10, or 15.

Available options: 5, 10, 15

prompt_extend · boolean · default: true

Enable intelligent prompt rewriting (slightly longer latency, better detail).

shot_type · enum<string> | null · default: single

Specifies the shot type for video generation.

Only takes effect when prompt_extend is enabled.

• single: Default value, outputs single-shot video
• multi: Outputs multi-shot video

Available options: single, multi

audio · boolean | null · default: true

Enable automatic audio generation. Set to false to force a silent output.

audio_url · string<uri> | null

Custom audio file URL (wav/mp3, 3-30s, ≤15MB). Overrides the audio flag.

seed · integer

Random seed [0, 2147483647].

Required range: 0 <= x <= 2147483647

Response

202 - application/json

Accepted - Task created successfully

task_info · object

Show child attributes