Welcome
Welcome to the documentation for Sapling.ai's developer platform. 🚀 Sapling is a platform for building language model applications. Sapling's API provides endpoints for text preprocessing, classification, prediction, correction, and more. Sapling's SDK provides convenient wrappers around the API, as well as UI components for quickly integrating LM-powered functionality in your web application.
Getting Started
Get started by following the instructions here. You can see some of these capabilities by visiting our Demo page.
From there, you can start using the Sapling HTTP API or SDK.
You can also try out Sapling using Postman or Zapier. For questions, email us or join our Discord:
Send HTTP requests | Create Zaps | Chat with us |
---|---|---|
Code Examples
Find example code in our Github repository, including examples for Python and JavaScript (React, Angular). Start a Github issue if your use case isn't handled.
Use Cases
Here are some of the things you can do with Sapling:
- ⚙️ Integrate deep learning-powered grammar/spell checking into any application, either through the SDK or through the HTTP endpoint.
- ✨ Autocomplete phrases and sentences to save users time typing.
- 😀 Check the tone of messages.
- Check the sentiment of messages.
- 🟰 Search for semantically similar terms to augment tagging and search.
- 🔍 Pull semantically relevant responses for knowledge-base retrieval and search.
- 💫 Rephrase sentences to adjust formality, length, or syntax.
- 🕵🏽♀️ Detect AI content or score the quality of content (contact us for details on quality scoring).
You can try out many of the examples on our utilities page.
New functionality is being released every week. Contact us to learn more or to discuss if Sapling can support your use case.
Key Principles
Sapling allows you to easily integrate language model-powered functionality into your application.
- End-to-end: Sapling is designed to handle the entire stack, from preprocessing text inputs, to model inference, to postprocessing outputs, to scoring and filtering outputs. It covers the last mile that's required for production systems.
- Predictable: Sapling is designed to not hallucinate or produce unexpected texts. It returns structured outputs that can reliably be used by production systems.
- Efficient: Sapling is designed for use in real-time settings. This is true for both Sapling's cloud API as well as its self-hosted offerings.
- Turnkey: For many applications, Sapling can be integrated with just a few lines of JavaScript/Python code. We ensure that the API/SDK are stable and well-documented (see our SLA).
- Secure: Whether you use our cloud and require finetuning or you require an airgapped deployment with zero data retention, Sapling can support your data security requirements.
Subscribing
Signing up and generating an API key will start you on a free trial of the API. There is a character limit and a 1-month expiration date for the trial.
Subscribe to remove quota limits and begin usage-based pricing (billed monthly). This will also increase the character limit (up to 10 million characters per day for the /edits
endpoint, as one example) and upgrade the key so that it does not expire.
About Us
Sapling is developed by former engineers and researchers from Berkeley, Google, Meta, and Stanford.