The primary purpose of this API is to provide word predictions and auto-correct suggestions to text entry interfaces. The API supports input via:

The API also supports a variety of other basic natural language processing (NLP) tasks, e.g. finding the most likely words in a sentence.

The documentation is organized into the following sections:

If you use this API in your research, please cite this paper:

@inproceedings{vertanen_velocitap,
  author       = {Keith Vertanen and Haythem Memmi and Justin Emge and Shyam Reyal and Per Ola Kristensson},
  title        = {VelociTap: investigating fast mobile text entry using sentence-based decoding of touchscreen keyboard input},
  booktitle    = {CHI '15: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems},
  location     = {Seoul, Korea},
  year         = {2015},
  pages        = {659--668},
}


API version history:

Version Date Description
2.6 Aug 2023 Added support for ambiguous keyboards and disambiguation of noisy input.
2.5 Jan 2021 Added distribution predictions.
2.4 Dec 2020 Deletion of keyboards after one hour of inactivity. Reorganization of documentation.
2.3 Sep 2020 Support for different decoder configurations and language models.
2.2 Sep 2020 Added ability to add keyboard, transform keyboard, and make predictions based on a tap sequence.
2.1 Apr 2020 Added future caching endpoints to mitigate latency impacts to user interfaces.
2.0 Mar 2020 Support for prefix, suffix, and substring in random word endpoint. Bug fixes to sort. Addition of character language model. Added log probability results to word prediction results.
1.0 Jun 2019 Initial version supporting basic natural language processing tasks. Designed to support mobile app development in a summer school for high school girls.

This material is based upon work supported by the National Science Foundation under Grant No. 1750193. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.