Pricing

AI Platform Data Labeling Service enables you to request human labeling for a collection of data that you plan to use to train a custom machine learning model. Prices for the service are computed based on:

  • The type of labeling task.
  • The number of annotation units generated. The unit for each type of task is described in the table in the Labeling costs section.
    • For an image classification task, units are determined by the label set size (for labeling quality concern, every 20 labels is a problem set and a price unit) and the number of human labelers. The price for single-label and multi-label classification is the same. For example, for an image, the label set size is 22, the number of human labelers is 3, this image will count for 2 * 3 = 6 units.
    • For an image bounding box task, units are the sum of labeler usages for each bounding box identified (labeler usages are related to the replication setting for this labeling job). Images without bounding boxes will not be charged. For example, if an image has two bounding boxes, with 3 labelers selected for each image, it will count for 2 * 3 = 6 units.
    • For an image segmentation task, units are the sum of labeler usages for each segmentation identified (labeler usages are related to the replication setting for this labeling job). Images without segmentations will not be charged. For example, if an image has two segmentations, with 3 labelers selected for each image, it will count for 2 * 3 = 6 units.
    • For an image rotated box task, units are the sum of labeler usages for each rotated box identified (labeler usages are related to the replication setting for this labeling job). Images without rotated boxes will not be charged. For example, if an image has two rotated boxes, with 3 labelers selected for each image, it will count for 2 * 3 = 6 units.
    • For an image polygon/polyline task, units are the sum of labeler usages for each polygon/polyline identified (labeler usages are related to the replication setting for this labeling job). Images without polygons/polylines will not be charged. For example, if an image has two polygons/polylines, with 3 labelers selected for each image, it will count for 2 * 3 = 6 units.
    • For a video classification task, units are determined by the video length (every 5 seconds is a price unit), label set size(for labeling quality concern, every 20 labels is a problem set and a price unit) and the number of human labelers. The price for single-label and multi-label classification is the same. For example, if a video is 20 seconds, the label set size is 22, the number of human labelers is 3, this video will count for 4 * 2 * 3 = 24 units.
    • For a video object detection task, unit are the sum of labeler usages for each bounding box identified (labeler usages are related to the replication setting for this labeling job). Video without objects will not be charged. For example, if one video has 2 objects, with 3 labelers selected for each video, it will count for 2 * 3 = 6 units.
    • For a video object tracking task, units are determined by the video length (every 30 seconds is a price unit) and the sum of labeler usages for each object identified(labeler usages are related to the replication setting for this labeling job). For example, if one video is 2 minutes, it has 2 objects, the number of human labelers is 3, it will count for 18 units (4 * 3 units for 2 minutes video with 3 labelers, 2 * 3 units for the objects with 3 labelers).
    • For a video event task, units are units are determined by the video length (every 30 seconds is a price unit) and the sum of labeler usages for each event identified(labeler usages are related to the replication setting for this labeling job). For example, if one video is 2 minutes, it has 2 events, the number of human labelers is 3, it will count for 18 units (4 * 3 units for 2 minutes video with 3 labelers, 2 * 3 units for the events with 3 labelers).
    • For a text classification task, units are determined by the words length (every 50 words is a price unit), label set size (for labeling quality concern, every 20 labels is a problem set and a price unit) and the number of human labelers. The price for single-label and multi-label classification is the same. For example, if one piece of text with 70 words is classified, the label set size is 22, the number of human labelers is 3. This text will count for 2 * 2 * 3 = 12 units.
    • For a text sentiment task, units are determined by the words length (every 50 words is a price unit), label set size(for labeling quality concern, every 20 labels is a problem set and a price unit) and the number of human labelers. The price for single-label and multi-label classification is the same. For example, if one piece of text with 70 words is analyzed, the label set size is 22, the number of human labelers is 3. This text will count for 2 * 2 * 3 = 12 units.
    • For a text entity extraction task, units are the sum of labeler usages for each entities identified (labeler usages are related to the replication setting for this labeling job). Text without entities will not be charged. For example, if one piece of text with 70 words is analyzed and 2 entities are extracted, with 3 labelers selected for each text, it will count for 2 * 3 = 6 units.
  • The number of human labelers who label each data item. This is specified by the HumanAnnotationConfig replica_count field in the API and the labelers per data item field when creating a new data labeling request in the Data Labeling Service UI. The number of units is multiplied by the number of labelers when calculating price. For example, if you request 3 human labelers for each data item, then you pay 3 times the cost.

Labeling costs

The table below provides the price per 1,000 units per human labeler, based on the unit listed for each objective. Tier 1 pricing applies to the first 50,000 units per month in each Google Cloud project; Tier 2 pricing applies to the next 950,000 units per month in the project, up to 1,000,000 units. Contact us for pricing above 1,000,000 units per month.

Data type Objective Unit Tier 1 Tier 2
Image Classification Image $35 $25
Bounding box Bounding box $63 $49
Segmentation Segment $870 $850
Rotated box Bounding box $86 $60
Polygon/polyline Polygon/Polyline $257 $180
Video Classification 5sec video $86 $60
Object detection Bounding box $86 $60
Object tracking Object in 30sec video $686 $480
Event Event in 30sec video $214 $150
Text Classification 50 words $129 $90
Sentiment 50 words $200 $140
Entity extraction Entity $86 $60

Google Cloud Platform costs

If you store data items to be labeled in Google Cloud Storage, or use other Google Cloud Platform resources in tandem with AI Platform Data Labeling Service, such as Google App Engine instances, then you will also be billed for the use of those services. See the Google Cloud Platform Pricing Calculator to determine other costs based on current rates.

To view your current billing status in the Cloud Console, including usage and your current bill, see the Billing page. For more details about managing your account, see the Cloud Billing Documentation or Billing and Payments Support.