Moondream3 Preview Detect

Moondream3 Preview Detect

Playground

Moondream3 Detect: Precise object bounding boxes in images for accurate computer vision localization. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

Moondream 3 Detect is a powerful vision-language model for identifying and localizing objects within images. It uses natural language input to detect specific items and returns their bounding box coordinates with high precision — ideal for visual search, annotation, and AI-assisted labeling.

✨ Key Features

• Natural Language Object Queries Simply describe what you want to detect — e.g., “person,” “car,” “dog,” “chair.”

• Accurate Bounding Boxes Returns precise x_min, y_min, x_max, y_max coordinates for each detected instance.

• Multi-Object Detection Supports multiple instances of the same category in one image.

• Fast and Lightweight Optimized for real-time or batch detection workflows with low latency.

⚙️ Example Usage

🔹 Detect Cars

{
 "image": "https://example.com/photo.jpg",
 "prompt": "car"
}

🔹 Detect People

{
 "image": "https://example.com/photo.jpg",
 "prompt": "person"
}

🔹 Detect Any Object

{
 "image": "https://example.com/photo.jpg",
 "prompt": "bicycle"
}

📦 Output Format

Bounding boxes are returned in normalized coordinates (range 0–1):

{
 "objects": [
 {
 "x_min": 0.1556,
 "x_max": 0.6881,
 "y_min": 0.2610,
 "y_max": 0.9551
 }
 ]
}

where

• (x_min, y_min) = top-left corner
• (x_max, y_max) = bottom-right corner

If multiple objects are detected, all boxes appear in the "objects" array.

💡 Best Practices

• Use specific, clear object names for best accuracy.
• For small or distant objects, higher-resolution images improve detection.
• Supported formats: JPEG, PNG, WebP
• Maximum image size: 10 MB

💰 Pricing

• $0.001 per request
• Contact WaveSpeedAI for bulk or enterprise pricing options.

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"
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/wavespeed-ai/moondream3-preview/detect" \
  -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