Skip to main content

API Pricing

Subscribe to the API here (using the email for your Sapling account):


note

This page provides pricing for only a subset of Sapling's endpoints. Please contact sales@sapling.ai for pricing for other endpoints.

If you need a self-hosted solution or a BAA for HIPAA compliance, please also contact sales@sapling.ai for custom pricing.

Usage Tiers


Languages Besides zh, jp, ko

The table below provides the costs for the Edits and Rephrase endpoints in terms of number of input characters to the API:

Usage Per MonthCost Per 1000 Characters
0-10 million chars$0.025
10 million-50 million chars$0.020
50 million-100 million chars$0.015
Over 100 million charsContact us

Pricing Calculator

1 million chars100 million chars
30 million chars = $650/month
10 million chars x $0.025/(1k chars) + 20 million chars x $0.02/(1k chars)

zh, jp, ko

To get the price per character for Chinese, Japanese, and Korean, multiply the above by 2.5. This is due to the larger size of the character sets for these languages, increasing the compute required per character.


  • Sapling also offers a spell check-only endpoint (does not include grammar correction) with lower pricing.
  • You'll be billed based on usage at the end of the monthly period.
  • After creating your API key, you can view your usage at https://sapling.ai/api_stats.

Calculating Usage

Sapling's pricing is per character. For English, the number of characters is ≈ the number of words × 5.

For example, suppose you need to process 500 articles per month. Each article is roughly 1000 words. That yields:

500 articles/month×1000 words/article×5 characters/word2,500,000 characters/month500 \text{ articles/month} \times 1000 \text{ words/article} \times 5 \text{ characters/word} \approx 2,500,000 \text{ characters/month}

You may be thinking: what if we need to send multiple revisions of a piece of text? Sapling caches suggestions for previously processed sentences. If you send a chunk of text, only modified, previously unprocessed sentences will be included in additional usage. Note, however, that even a single character change to a sentence can turn it into a previously unseen sentence.

Example 1

Suppose you process this text (containing 83 characters):

Alice followed the rabbit. She saw a blue caterpillar. She went to a mad tea party.

You then modify the text and send the following instead

Alice followed the rabbit. She was questioned by a caterpillar. She went to a mad tea party.

Note that the 1st and 3rd sentences are unchanged, and hence won't count towards usage. However, the 2nd sentence was modified. Even though the original 2nd sentence and the modified 2nd sentence both contain She, a, and caterpillar, Sapling will count all the characters in the new sentence. Hence, the total usage will be 83 + 36 = 119 characters.

Example 2

Suppose you use Sapling's SDK with a text editor on your website. You enter the 83 character text above.

You then refresh the page and the saved text (still 83 characters) in the editor is sent to Sapling again.

The total usage in this case will be 83 characters since the text sent is completely unchanged.

API vs. Frontend Integrations

If your team plans to use Sapling with an application where a Sapling integration does not exist, then the API or SDK is the only option. However, there are many platforms where one of Sapling's integrations can be used. You can find a list of our integrations here.

For some use cases, the frontend integrations may provide functionality that is not available in the API or SDK. For example, if you use snippet macros, the Sapling browser integrations provide search and text expansion functionality not currently offered in the API.

Benefits of the API include customizability and configurability. Benefits of the frontend integration include additional functionality and immediate use of a well-tested interface. Another consideration is, of course, the development time to create and maintain the API integration.

Lastly, there is the price difference. Usage of Sapling via frontend integrations is priced per seat. If you have users who each process large volumes of text, seat-based pricing may be lower.

Quota Limits vs. Metered Pricing

The tables above show metered pricing for different usage tiers.

For enterprise teams with high but predictable usage volumes, Sapling offers discounted pricing. You can purchase an upfront quota for an agreed upon time window (e.g., a monthly or annual quota) at a discounted price. Usage above that quota will be pro-rated. However, usage below that quota will not carry over past the time window.

We suggest that high-volume users consider a baseline quota for discounted pricing.

Tracking and Managing Usage

After creating your API key, you can view your usage at https://sapling.ai/api_stats. It may take up to 6 hours for the usage to update. The daily usage breakdown should also be provided with your receipt from Stripe.

If your usage is higher than expected, here are a few recommendations:

  1. (If using API) Store your API call history in a simple database/log. Sometimes higher than expected usage is due to spiky usage or a flurry of usage over a short period of time.
  2. (If using SDK) Try using the checkOnce function to control how frequently requests are sent to the Sapling endpoint.
  3. Consider only activating Sapling for a subset of the editable elements or API calls. For example, if you have a team of both writers and editors, it may be more important that editors can access Sapling's suggestions.
  4. Ask the Sapling team to throttle usage to a certain limit per month. You'll need to specify whether you want the throttling to evenly distribute usage throughout the month (default) or allow spiky/frontloaded usage.

Contact us if you need more fine-grained tracking of usage.