Einstein Language
Einstein Language is a retired set of three Salesforce natural language APIs: Einstein Sentiment, Einstein Intent, and Einstein NER.
Definition
Einstein Language is a retired set of three Salesforce natural language APIs: Einstein Sentiment, Einstein Intent, and Einstein NER. It ran on the api.einstein.ai host under Einstein Platform Services, beside Einstein Vision, and authorized with a JWT signed by a private key you downloaded.
Salesforce announced the retirement in 2024, and the developer guide still carries the banner. Subscriptions stopped being renewable after May 1, 2024, and orgs had to move before their existing Order Term ran out.
There was no single shutdown date and no one-for-one successor. Each of the three jobs got absorbed by the product that needed it.
In plain English
“Salesforce used to rent out three text-reading services. One guessed the mood of a sentence. The second sorted sentences into buckets you defined. Dates, money, and names were the third one's job. Salesforce stopped selling them.”
Worked example
A support team at a mid-size subscription company trained an Einstein Intent model in 2023. They uploaded a CSV of 150 labeled utterances across five labels: Billing, Order Change, Password Help, Sales Opportunity, and Shipping Info. A POST to https://api.einstein.ai/v2/language/datasets/upload with type=text-intent returned a dataset ID, and a call to /v2/language/train returned a model ID. Once training read SUCCEEDED, an Apex callout sent inbound Case Description text to /v2/language/intent, took the top-ranked label, and wrote it to a Predicted_Intent__c picklist that drove assignment rules. When the subscription lapsed, the callout failed and the team rebuilt that routing inside an Einstein Bot intent set.
What Einstein Language shipped, and where its three jobs went
The three APIs and what each returned
Sentiment scored a snippet positive, negative, or neutral and shipped a prebuilt model called CommunitySentiment, trained on community posts, Chatter, and reviews, usable with just a token. Intent demanded your own labeled CSV, an async training run, and a model ID on every call. NER never left beta, and tagged eleven fixed types including DATETIME, MONEY, ORGANIZATION, PERSON, and PHONE-NUMBER.
Authorization put people off
These calls never touched your org session. You downloaded a private key named einstein_platform.pem, signed an assertion with it, and traded that for a bearer token with an expiry you chose. Nothing refreshed itself, so every consumer needed renewal logic and a safe place for the key. For an admin with no secrets store, that was a wall.
Why Salesforce pulled it
A general-purpose text API outside the CRM data model was always an awkward fit. Features needing intent classification grew their own: bot intent models train inside Bot Builder on utterances an admin can read and correct. Then large language models made a fixed eleven-type extractor look thin next to a prompt returning any shape you ask for.
Where each job lives now
Map by use case, not by endpoint. Conversational intent goes to Einstein Bot intent sets. Record-level classification is Einstein Case Classification, not Einstein Prediction Builder, which is on the retirement list too. Service sentiment ships inside Service Cloud. Extraction and summarization become a Prompt Builder template with a defined output, under the Einstein Trust Layer. Run the new path beside the old logic first, because label boundaries rarely line up.
Trust & references
Cross-checked against the following references.
- Introduction to Salesforce Einstein Language (opens in new tab)Salesforce
- What Is Einstein NER (Beta)? (opens in new tab)Salesforce
- Einstein Intent Quick Start (opens in new tab)Salesforce
- Community Sentiment Model (opens in new tab)Salesforce
- Salesforce Active Product & Feature Retirements (opens in new tab)Salesforce
Straight from the source - Salesforce's reference material on Einstein Language.
- Set Up Authorization (opens in new tab)Salesforce
- Supported Languages (opens in new tab)Salesforce
Hands-on resources to go deeper on Einstein Language.
About the Author
Dipojjal Chakrabarti is a B2C Solution Architect with 29 Salesforce certifications and over 13 years in the Salesforce ecosystem. He writes and edits salesforcedictionary.com, published by KineticBit Inc., to help admins, developers, architects, and cert/interview candidates sharpen their fundamentals. More about Dipojjal.
Test your knowledge
+5 pts / dayQ1. How did Salesforce frame the retirement of Einstein Language?
Q2. Which Einstein Language API never made it out of beta?
Q3. You inherit an Apex callout that posts Case Description text to the Einstein Intent endpoint and writes the top label to a picklist. What is the closest supported replacement?
Discussion
Loading discussion…