Patina Material Extract

Patina Material Extract

Playground

PATINA Material Extract turns any photograph or reference image into a complete seamlessly tiling PBR material set (basecolor, normal, roughness, metalness, height), guided by a text prompt. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

Patina Material Extract extracts a seamlessly tiling PBR material from any reference image. Upload a photo of a real-world surface, describe which texture to extract, and the model produces a complete tileable material map set — isolating the target surface from complex scenes and converting it into production-ready PBR maps.

• Just have a plain surface photo? Try Patina Image-to-Map for direct map generation.
• Need to generate from text? Try Patina Material for prompt-driven PBR creation.

Why Choose This?

• Surface extraction from complex scenes Isolate and extract a specific material from photos that contain multiple surfaces, objects, or backgrounds — guided by your text description.

• Seamlessly tiling output Produces tileable material maps ready for use on any geometry without visible seams.

• Tiling direction control Choose omnidirectional tiling (both), horizontal-only, or vertical-only to match your UV mapping needs.

• Custom output size Specify any output resolution to match your target quality and performance requirements.

• Production-ready PBR maps Output is formatted for direct use in game engines (Unreal, Unity), 3D tools (Blender, Maya), and real-time rendering pipelines.

Parameters

How to Use

1. Upload your reference image — provide a photo containing the surface you want to extract.
2. Write your prompt — describe which specific texture to extract from the image (e.g. “the brick wall in the background”, “the leather seat material”).
3. Set size (optional) — specify output dimensions to match your target resolution.
4. Choose tiling_mode (optional) — select both for omnidirectional tiling, or horizontal/vertical for axis-specific tiling.
5. Submit — the model extracts and generates all PBR material maps in a single run.
6. Download your complete tileable map set ready for use in your 3D pipeline.

Pricing

Just $1.02 per run (6 images).

Best Use Cases

• Game development — Extract tileable PBR materials directly from photo references for environment and prop texturing.
• Architectural visualization — Digitize real-world surface materials from photography for accurate scene reproduction.
• Asset digitization — Convert complex real-world scene photos into individual, reusable tileable material assets.
• 3D environment art — Extract specific surface materials from concept art, mood boards, or reference photography.
• Texture library building — Rapidly generate large libraries of tileable PBR materials from photo collections.

Pro Tips

• Be specific in your prompt about which surface to extract — include location, color, or texture characteristics (e.g. “the dark grey concrete floor” rather than just “concrete”).
• Use images where the target surface occupies a clearly visible area for the most accurate extraction.
• Avoid heavily shadowed or overexposed areas of the target surface for cleaner albedo map results.
• Use tiling_mode = both for floor, wall, and terrain materials; use horizontal or vertical for directional surfaces like planks or fabric.

Notes

• Both image and prompt are required fields.
• Each run generates a full set of PBR map types as output.
• Ensure image URLs are publicly accessible if using a link rather than a direct upload.
• Please ensure your content complies with WaveSpeed AI’s usage policies.

Related Models

• Patina Image-to-Map — Generate PBR maps directly from a plain surface photograph.
• Patina Material — Generate tileable PBR materials from a text description.

Authentication

For authentication details, please refer to the Authentication Guide.

API Endpoints

Submit Task & Query Result

set -euo pipefail

export WAVESPEED_API_KEY="your-api-key"

REQUEST_BODY=$(cat <<'JSON'
{
  "prompt": "A cinematic ocean wave at sunrise, highly detailed",
  "image": "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg",
  "size": "1024*1024",
  "tiling_mode": "both"
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/wavespeed-ai/patina/material-extract" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}")

TASK=$(printf '%s' "${SUBMIT_RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "${TASK}" | jq -r '.id // empty')
if [ -z "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id
' >&2
  exit 1
fi
RESULT_URL=$(printf '%s' "${TASK}" | jq -r '.urls.get // empty')
if [ -z "${RESULT_URL}" ]; then RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"; fi

# 2. Poll until the prediction finishes.
while true; do
  RESPONSE=$(curl --silent --show-error --fail-with-body \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}")
  RESULT=$(printf '%s' "${RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

  case "${STATUS}" in
    completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
    failed|cancelled|timeout) printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    created|processing) sleep 2 ;;
    *) printf 'Unexpected status: %s
' "${STATUS}" >&2; exit 1 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Response Parameters

Result Request Parameters

Result Response Parameters