Language pricing will be changing on tasks created starting October 10th. Language pricing will now depend on desired fluency level. These will be applied as multipliers to the base task cost per the following table:
No, there is no minimum for using the self-labeling option.
You get 200 free labeling units every month. One annotation is defined as one labeling unit, and one attribute is equivalent to 0.5 labeling units. If you exceed 200 units, each additional unit will cost 5 cents.
There is no limit to the number of users! We would caution, however, that there are no features on self-label that allow for permissioning. All users have access to the same features - including project set up. As such, we recommend limiting the number of self-label users to just those who are close to the project
Yes. A common workflow is self-labeling some batches to create strong quality tasks and then sending regular batches to Scale’s workforce. Additionally, you can send the remainder of tasks in a self-label batch to Scale’s workforce if you choose.
Yes! Scale Studio allows you to bring your own team of labelers to the labeling platform. Studio allows you to create training courses and monitor your labeler’s performance over time. Learn more here.
All of Rapid’s task types have self-labeling support, including image, video, text, lidar, document, and audio tasks.
Yes. When creating a batch, be sure to include the self_label_batch flag. You will then be able to self-label any tasks associated with this batch.