Prodigy for Research Institutions

Run experiments that ask new questions, and share datasets that lead to convincing answers. Researchers use Prodigy and our open-source library spaCy in fields such as digital humanties, bioinformatics, and machine learning.


universities and research institutions
  • special site-wide license for degree-granting instutions
  • 12 months unlimited usage for all students and staff
  • unlimited annotators
  • Prodigy installer, web application and extensive documentation
$10,000 USD
per year, site-wide (excl. tax)


interim license
  • available to researchers at degree-granting academic institutions only
  • apply for a free 3-month license and keep working
  • receive details to help your institution get set up with a site-wide license

What’s included

Need more info? Try the Prodigy live demo, or see the Frequently Asked Questions and Terms and Conditions.
If you're a freelancer or working for a company, see this page on professional licenses for Prodigy.

How Prodigy can help researchers

We hope Prodigy can help the research community to conduct rigorous evaluations of a wider variety of fine-grained, incisive questions. It's a simple and efficient tool for any type of data coding, and can be used to train and evaluate custom statistical models for specific hypotheses. For machine learning researchers, Prodigy is particularly effective hard-to-evaluate tasks such as translation, synthesis and generation. With Prodigy, you can quickly perform randomized A/B tests, export the decisions, and attach them to your publication, giving a repeatable evaluation.

To Close Digital Divide, Microsoft to Harness Unused Television Channels
Microsoft is buying television bandwidth
Microsoft is selling its shares in local news stations
Photo of a pug wrapped in a blanket outdoors
by: Matthew Henry url:
 car  car  car  car 
by: Nabeel Syed url:

One of the best tools I came across for textual annotation, made by the great minds behind spaCy.

Uri Goren – Research Engineer, Yahoo

Thanks for the great app. I literally downloaded it at breakfast and validated an idea by lunch time.

Lev Konstantinovskiy – NLP Engineer, FullFact

I'm a huge fan of everything Explosion AI does. Their new annotation tool is the best I've ever seen.

A lack of labeled data held geoparsing back for years. It took a week to fix that with Prodigy.

Andy Halterman – Researcher, MIT

Citing Prodigy

We're preparing a short technical report for Prodigy, which will be uploaded to arXiv. You can cite the report as follows: "Prodigy: A new annotation tool for radically efficient machine teaching", Ines Montani and Matthew Honnibal, 2017 (to appear).

    author    = "Montani, Ines and Honnibal, Matthew",
    title     = "Prodigy: A new annotation tool for radically efficient machine teaching",
    journal   = "Artificial Intelligence",
    volume    = "to appear",
    year      = "2018",
    eprint    = "to appear"