3D Generation

Three endpoints turn an image or a text prompt into a 3D model. All three return { "jobId": "..." } immediately — see Jobs for how to poll and download the result. For pricing, see Pricing & Credits.

Text-to-3D

POST /jobs/text-to-3d
Content-Type: application/json

Field	Type	Required	Default	Notes
prompt	string	yes	—	Up to 4096 characters.
topology	"auto" | "quads" | "triangles"	no	"auto"	Mesh topology of the output.
targetFaceCount	integer	no	0	0 lets the pipeline pick. Max 500000.
textureResolution	"128" | "256" | "512" | "1k" | "2k" | "4k" | "auto" | "none"	no	"auto"	"none" produces a geometry-only model (no texture).
tPose	boolean	no	false	Generate the subject in a neutral T-pose — useful if you plan to rig/animate it afterward.
apiVersion	integer	no	1	Only 1 exists today.

curl https://api.sloyd.ai/api/jobs/text-to-3d \
  -H "x-client-id: $SLOYD_CLIENT_ID" \
  -H "x-client-secret: $SLOYD_CLIENT_SECRET" \
  -H "Content-Type: application/json" \
  -d '{ "prompt": "a low-poly pine tree", "textureResolution": "2k" }'

Image-to-3D

POST /jobs/image-to-3d
Content-Type: multipart/form-data

The image is a file upload, not a JSON field.

Field	Type	Required	Default	Notes
file	image file	yes*	—	PNG/JPEG etc. Max 20MB. *Not required if imageJobId is given (see Uploading an image for reuse).
targetFaceCount	integer	no	0	Same as text-to-3d.
textureResolution	enum	no	"auto"	Same values as text-to-3d.
topology	enum	no	"auto"	Same values as text-to-3d.
apiVersion	integer	no	1

curl https://api.sloyd.ai/api/jobs/image-to-3d \
  -H "x-client-id: $SLOYD_CLIENT_ID" \
  -H "x-client-secret: $SLOYD_CLIENT_SECRET" \
  -F "file=@helmet.png" \
  -F "targetFaceCount=20000"

Multi-Image-to-3D

POST /jobs/multi-image-to-3d
Content-Type: multipart/form-data

Generates from up to four views of the same subject. frontImage is required; at least one of leftImage / backImage / rightImage is also required.

Field	Type	Required	Notes
frontImage	image file	yes	Max 20MB.
leftImage / backImage / rightImage	image file	at least one	Max 20MB each.
FaceCount	numeric string	no	e.g. "40000".
GenerateType	"Normal" | "LowPoly" | "Geometry" | "Sketch"	no
PolygonType	"triangle" | "quadrilateral"	no
EnablePBR	"true" | "false"	no	String, not boolean.
ModelNo	"3.0" | "3.1"	no	Underlying model version.
apiVersion	integer	no	Default 1.

curl https://api.sloyd.ai/api/jobs/multi-image-to-3d \
  -H "x-client-id: $SLOYD_CLIENT_ID" \
  -H "x-client-secret: $SLOYD_CLIENT_SECRET" \
  -F "frontImage=@front.png" \
  -F "leftImage=@left.png" \
  -F "backImage=@back.png"

Uploading an image for reuse

POST /jobs/image-upload
Content-Type: multipart/form-data

Uploads a standalone image and returns its own jobId. Pass that id as imageJobId to image-to-3d to reuse an already-uploaded image without uploading the bytes again. Free (0 credits).

Field	Type	Required
file	image file	yes

Response: { "jobId": "...", "metaJson": {...} }.