MODEL ID microsoft:trellis-2@4b

live

TRELLIS.2

by Microsoft December 17, 2025

TRELLIS.2 is a 4-billion-parameter 3D generative model that transforms 2D images into fully textured 3D assets with complex topology, sharp features, and physically based rendering materials. It uses a novel field-free sparse voxel representation and scalable structured latent space to enable efficient, high-resolution 3D generation from single image input.

API Options

Platform-level options for task execution and delivery.

taskType string required value: 3dInference: Identifier for the type of task being performed

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

outputType string default: URL: 3D output type.

Allowed values 1 value

outputFormat string default: GLB

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

`GLB`: Binary glTF format for 3D models, including geometry, textures, and PBR materials.

Allowed values 1 value

webhookURL string URI

Specifies a webhook URL where JSON responses will be sent 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
PLATFORM

deliveryMethod string default: async

Determines how the API delivers task results.

Allowed values 2 values

: Returns complete results directly in the API response.
: Returns an immediate acknowledgment with the task UUID. Poll for results using getResponse.

Learn more 1 resource

Task Polling
PLATFORM

uploadEndpoint string 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.

// 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.

ttl integer min: 60: Time-to-live (TTL) in seconds for generated content. Only applies when outputType is URL.

includeCost boolean default: false: Include task cost in the response.

numberResults 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.

inputs » image image string required: Image input (UUID, URL, Data URI, or Base64).

Generation Parameters

Core parameters for controlling the generated content.

model string required value: microsoft:trellis-2@4b

Identifier of the model to use for generation.

Learn more 3 resources

seed integer min: 0 max: 9223372036854776000

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

Learn more 1 resource

Text To Image: Seed Controlling Randomness Deterministically
GUIDE

Settings

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

settings » resolution resolution integer default: 1024: Voxel grid resolution for 3D generation. Higher values capture finer geometric detail.

Allowed values 3 values

settings » decimationTarget decimationTarget integer min: 100000 max: 1000000 default: 500000: Target polygon count for mesh simplification. Lower values produce lighter models at the cost of geometric detail.

settings » remesh remesh boolean default: true: Whether to remesh the output for cleaner topology.

settings » shapeSlat shapeSlat object

Sampler for the Structured Latent stage that refines the 3D geometry.

Properties 4 properties

settings » shapeSlat » steps steps integer min: 1 max: 50: Number of denoising steps. More steps improve quality but increase generation time.

settings » shapeSlat » guidanceRescale guidanceRescale float min: 0 max: 1: Dampens oversaturation from high guidance values. Lower values apply more correction.

settings » shapeSlat » guidanceStrength guidanceStrength float min: 1 max: 10: How closely the output follows the input image. Higher values enforce stricter adherence to the visible geometry.

settings » shapeSlat » rescaleT rescaleT float min: 1 max: 6: Timestep rescaling factor for the sampling schedule. Controls how guidance strength is distributed across steps.

settings » sparseStructure sparseStructure object

Sampler for the initial sparse voxel grid that establishes the basic 3D geometry.

Properties 4 properties

settings » sparseStructure » steps steps integer min: 1 max: 50: Number of denoising steps. More steps improve quality but increase generation time.

settings » sparseStructure » guidanceRescale guidanceRescale float min: 0 max: 1: Dampens oversaturation from high guidance values. Lower values apply more correction.

settings » sparseStructure » guidanceStrength guidanceStrength float min: 1 max: 10: How closely the output follows the input image. Higher values enforce stricter adherence to the visible geometry.

settings » sparseStructure » rescaleT rescaleT float min: 1 max: 6: Timestep rescaling factor for the sampling schedule. Controls how guidance strength is distributed across steps.

settings » texSlat texSlat object

Sampler for the Structured Latent stage that generates surface appearance.

Properties 4 properties

settings » texSlat » steps steps integer min: 1 max: 50: Number of denoising steps. More steps improve quality but increase generation time.

settings » texSlat » guidanceRescale guidanceRescale float min: 0 max: 1: Dampens oversaturation from high guidance values. Lower values apply more correction.

settings » texSlat » guidanceStrength guidanceStrength float min: 1 max: 10: How closely the output follows the input image. Higher values enforce stricter adherence to the visible geometry.

settings » texSlat » rescaleT rescaleT float min: 1 max: 6: Timestep rescaling factor for the sampling schedule. Controls how guidance strength is distributed across steps.

settings » textureSize textureSize integer min: 1024 max: 4096 step: 1024 default: 2048: Resolution of the output texture map in pixels. Higher values produce finer surface detail.

taskType string required value: 3dInference: Type of the task.

taskUUID string required UUID v4: UUID of the task.

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

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

modelUUID string required UUID v4: UUID of the output 3D model file.

modelURL string URI: URL of the output 3D model file.

modelBase64Data string: Base64-encoded 3D model file data.

modelDataURI string URI: Data URI of the output 3D model file.

Image to 3D

Orc Character Bust

Request

{
  "taskType": "3dInference",
  "taskUUID": "bd373968-8512-4315-a95f-95a4401fea3e",
  "model": "microsoft:trellis-2@4b",
  "seed": 424242,
  "settings": {
    "textureSize": 2048,
    "decimationTarget": 500000,
    "remesh": true,
    "resolution": 1024,
    "sparseStructure": {
      "guidanceStrength": 8,
      "guidanceRescale": 0.7,
      "steps": 30,
      "rescaleT": 3
    },
    "shapeSlat": {
      "guidanceStrength": 7.5,
      "guidanceRescale": 0.7,
      "steps": 28,
      "rescaleT": 3
    },
    "texSlat": {
      "guidanceStrength": 8.5,
      "guidanceRescale": 0.6,
      "steps": 30,
      "rescaleT": 3
    }
  },
  "inputs": {
    "image": "https://assets.runware.ai/assets/inputs/41d320a9-1ed2-4dd8-9727-d83297535d02.jpg"
  }
}

Response

{
  "taskType": "3dInference",
  "taskUUID": "bd373968-8512-4315-a95f-95a4401fea3e",
  "seed": 424242,
  "cost": 0.055,
  "outputs": {
    "files": [
      {
        "uuid": "112a24e5-7976-46b3-a22c-e658daaffdfe",
        "url": "https://im.runware.ai/image/os/a01d21/ws/4/ii/112a24e5-7976-46b3-a22c-e658daaffdfe.glb"
      }
    ]
  }
}

Image to 3D

Ornate Dragon Figurine

Request

{
  "taskType": "3dInference",
  "taskUUID": "5efab433-316e-4b0b-b8c4-6e0edbe3779c",
  "model": "microsoft:trellis-2@4b",
  "seed": 246813579,
  "settings": {
    "textureSize": 4096,
    "decimationTarget": 500000,
    "remesh": true,
    "resolution": 1536,
    "sparseStructure": {
      "guidanceStrength": 8,
      "guidanceRescale": 0.7,
      "steps": 30,
      "rescaleT": 3
    },
    "shapeSlat": {
      "guidanceStrength": 7.5,
      "guidanceRescale": 0.7,
      "steps": 28,
      "rescaleT": 3
    },
    "texSlat": {
      "guidanceStrength": 8.5,
      "guidanceRescale": 0.6,
      "steps": 32,
      "rescaleT": 3
    }
  },
  "inputs": {
    "image": "https://assets.runware.ai/assets/inputs/59b5b9e2-7083-4adf-88be-2793dab06233.jpg"
  }
}

Response

{
  "taskType": "3dInference",
  "taskUUID": "5efab433-316e-4b0b-b8c4-6e0edbe3779c",
  "seed": 246813579,
  "cost": 0.0198,
  "outputs": {
    "files": [
      {
        "uuid": "2dba7e00-adc6-410e-bcf9-7011266bf24f",
        "url": "https://im.runware.ai/image/os/a24d12/ws/4/ii/2dba7e00-adc6-410e-bcf9-7011266bf24f.glb"
      }
    ]
  }
}

Image to 3D

Ornate Fantasy Helmet

Request

{
  "taskType": "3dInference",
  "taskUUID": "0ccf2b8e-db45-4439-bba8-6a1420848399",
  "model": "microsoft:trellis-2@4b",
  "seed": 424242,
  "settings": {
    "textureSize": 4096,
    "decimationTarget": 500000,
    "remesh": true,
    "resolution": 1536,
    "sparseStructure": {
      "guidanceStrength": 8,
      "guidanceRescale": 0.7,
      "steps": 30,
      "rescaleT": 3
    },
    "shapeSlat": {
      "guidanceStrength": 7.5,
      "guidanceRescale": 0.7,
      "steps": 28,
      "rescaleT": 3
    },
    "texSlat": {
      "guidanceStrength": 8.5,
      "guidanceRescale": 0.6,
      "steps": 32,
      "rescaleT": 3
    }
  },
  "inputs": {
    "image": "https://assets.runware.ai/assets/inputs/84e77cec-bda0-4930-94f8-cacad245251d.jpg"
  }
}

Response

{
  "taskType": "3dInference",
  "taskUUID": "0ccf2b8e-db45-4439-bba8-6a1420848399",
  "seed": 424242,
  "cost": 0.0454,
  "outputs": {
    "files": [
      {
        "uuid": "7e34d8eb-0823-4d73-bde7-8599a08bb505",
        "url": "https://im.runware.ai/image/os/a05d22/ws/4/ii/7e34d8eb-0823-4d73-bde7-8599a08bb505.glb"
      }
    ]
  }
}

Image to 3D

Hero Product Figurine

Request

{
  "taskType": "3dInference",
  "taskUUID": "a8970b6f-a3d2-4fcd-abdb-6d44ac3477f0",
  "model": "microsoft:trellis-2@4b",
  "seed": 424242,
  "settings": {
    "textureSize": 4096,
    "decimationTarget": 500000,
    "remesh": true,
    "resolution": 1536,
    "sparseStructure": {
      "guidanceStrength": 8,
      "guidanceRescale": 0.7,
      "steps": 28,
      "rescaleT": 3
    },
    "shapeSlat": {
      "guidanceStrength": 7.5,
      "guidanceRescale": 0.7,
      "steps": 30,
      "rescaleT": 3
    },
    "texSlat": {
      "guidanceStrength": 8.5,
      "guidanceRescale": 0.6,
      "steps": 32,
      "rescaleT": 3
    }
  },
  "inputs": {
    "image": "https://assets.runware.ai/assets/inputs/223fe72c-93a2-4a97-9c2a-81183a365353.jpg"
  }
}

Response

{
  "taskType": "3dInference",
  "taskUUID": "a8970b6f-a3d2-4fcd-abdb-6d44ac3477f0",
  "seed": 424242,
  "cost": 0.0314,
  "outputs": {
    "files": [
      {
        "uuid": "6568996e-9567-4a40-84b4-1cc6202fe043",
        "url": "https://im.runware.ai/image/os/a03d21/ws/4/ii/6568996e-9567-4a40-84b4-1cc6202fe043.glb"
      }
    ]
  }
}