Abstract amber curves on a black background

Photo: Pawel Czerwinski / Unsplash

Guide

Einstein vs. Agentforce: what changed and what you already own

How Salesforce's Einstein features map to Agentforce names today, which ones were renamed, replaced or left alone, what that means for your org, and how to avoid buying the same AI twice.

Einstein is the brand Salesforce has used for AI since 2016, mostly for predictive features such as lead scoring, activity capture and case classification. Agentforce is the agent platform it launched in 2024. Some Einstein products were renamed, such as Einstein Copilot, which is now Agentforce. Einstein Bots was replaced and is in maintenance mode. Many predictive features still carry the Einstein name. Before you buy anything new, map what your contract already covers.

Why is Salesforce AI naming so confusing right now?

Because three things changed at once: product names, the bundles they are sold in, and the platform branding around them. Old and new names now sit side by side in Setup, in help articles and on order forms.

Salesforce did not retire Einstein in one move. It renamed some products, launched replacements for others and left many alone. Its help site still files articles under a section called Agentforce and Einstein Generative AI. A buyer reading a renewal quote can easily assume two names mean two products, or that one name covers everything. Neither assumption is safe.

What did the Einstein name cover originally?

Einstein started as predictive AI built into each Salesforce cloud. Generative and conversational tools arrived later under the same brand, which is where most of the confusion began.

Salesforce introduced Einstein on September 19, 2016. The launch release described AI embedded across every cloud and trained on customer, activity and other Salesforce data. Sales Cloud Einstein included predictive lead scoring, opportunity insights and automated activity capture. Service Cloud Einstein offered recommended case classification, recommended responses and predicted close times. Marketing Cloud Einstein added predictive scoring, predictive audiences and send-time optimization.

The generative era came next. Salesforce announced Einstein GPT on March 7, 2023, describing it as generative AI for CRM. On September 12, 2023 it announced Einstein Copilot, a conversational assistant, alongside the Einstein 1 Platform. In March 2024 it launched Einstein 1 Studio, grouping Copilot Builder, Prompt Builder and Model Builder. By then, Einstein meant scoring models, chat bots, a copilot and a builder toolkit, all at once.

When did Agentforce take over, and which names actually changed?

Salesforce unveiled Agentforce on September 12, 2024 and announced general availability on October 29, 2024. The renames and repackaging followed in stages over the next two years.

  • Einstein Copilot became Agentforce. The Winter '25 release notes say the agent type was renamed with no change in functionality, available starting January 2025.
  • Einstein Service Agent, announced in July 2024, is now Agentforce Service Agent, according to an editor's note on Salesforce's own announcement.
  • Agentforce add-ons and Agentforce 1 Editions became generally available on June 17, 2025. Salesforce said these replace the Einstein add-ons and Einstein 1 Editions.
  • From June 17, 2025, Agentforce (Default), the renamed copilot, stopped receiving new features. Salesforce recommends migrating to the Agentforce Employee agent.
  • Agentforce 360 was announced on October 13, 2025. Salesforce's help documentation says Data Cloud was rebranded to Data 360 as of October 14, 2025.
  • Einstein Studio is now called AI Models in Salesforce's developer documentation.
  • Salesforce's newest packaging, dated September 3, 2026, sells Core, Advanced and Max editions of Agentforce Sales, Agentforce Service and Agentforce Industries.

Some names did not change. Einstein Lead Scoring, Einstein Activity Capture and the Einstein Trust Layer all keep their names in current help articles. Einstein Bots also keeps its name, but its status is different, as the table shows.

How do the old Einstein names map to what Salesforce sells today?

Use this table as a translation guide when you read a contract, a quote or a Setup menu. The licensing column is a prompt for questions, not a statement of what your org owns.

Einstein product names and their current Salesforce equivalents
Name you may seeCurrent name or statusWhat it doesLicensing note to confirm
Einstein Lead ScoringStill Einstein Lead Scoring, documented under Sales Cloud EinsteinScores leads against your own past conversion patternsHelp lists it with Sales Cloud Einstein (included in legacy Performance and Unlimited, extra cost in Enterprise); on Core, Advanced and Max, confirm whether it is included or an add-on
Einstein Activity CaptureStill Einstein Activity CaptureSyncs Microsoft or Google email and calendar activity to related recordsReachable through several editions and add-ons, so check which one actually grants it
Einstein BotsStill Einstein Bots, in maintenance mode from August 17, 2026Scripted chat bots built from dialogs and intentsExisting bots stay supported; new agent work points to Agentforce Service Agent
Einstein CopilotAgentforce (Default), with Agentforce Employee agent as the recommended successorAssistant that helps employees inside SalesforceAsk how your current assistant entitlement maps to employee agents
Einstein Service AgentAgentforce Service AgentCustomer-facing agent for service conversationsConfirm which consumption or per-user model covers it
Einstein GPTNo single rename notice found; generative features now sit under AgentforceGenerated emails, replies and summariesAsk which current SKU covers each generative feature you use
Einstein Studio (Model Builder)AI ModelsConfigures foundation models and supports bringing your own LLMConfirm whether it comes with Data 360 or another entitlement
Einstein 1 Editions and Einstein add-onsAgentforce 1 Editions and Agentforce add-ons; newer Core, Advanced and Max editionsBundles of AI features with a CRM editionAsk whether renewal converts the old SKU or adds a new one
Einstein Trust LayerStill Einstein Trust LayerMasking, grounding, toxicity checks and audit for generative and agent featuresComes with qualifying generative AI entitlements rather than as its own purchase
Data CloudData 360Unified data layer used to ground agentsSame functionality, new name; check credit allocations on your order form

Will the Einstein features already in my org keep working?

In most cases, yes. A rename does not switch a feature off, and Salesforce described the copilot rename as having no functional change. The exceptions are features placed in maintenance mode, which keep running but stop improving.

Expect to see both names for a while. The Einstein Trust Layer help article still lists Einstein add-ons in its required editions. The permission set that grants access to Agentforce (Default) still carries a Copilot name. Admins searching Setup for one name may need to try the other.

Two features deserve a closer look. Einstein Bots continue to be supported after August 17, 2026, but receive no new features. Salesforce offers an Upgrade Agent tool that builds a separate Agentforce Service Agent from a bot's dialogs, intents and action steps. The original bot keeps running while you review the new one. Agentforce (Default) is similar. Salesforce advises finishing the move to an Employee agent before any related license change, to avoid gaps in availability.

How can I avoid paying twice for the same AI capability?

Inventory what your contract already grants before you accept any new AI line item. Overlap usually comes from one feature being reachable through several bundles.

Einstein Activity Capture shows the problem clearly. Its help article lists a standard version in several editions. It also lists inclusion in Unlimited, Einstein 1 Sales and Agentforce 1 Editions. Einstein for Sales, Sales Engagement and Revenue Intelligence each provide it as well. An org that bought one of those add-ons for activity sync may already have it from its edition.

The same check applies at renewal. Salesforce says Agentforce add-ons replace Einstein add-ons. Its newest editions announcement adds that Agentforce 1 customers can move to Max at no extra cost. Neither statement tells you how your own order form will be rewritten. Ask for the old and new SKUs side by side.

  • List every AI-related SKU on your current order form, including retired Einstein names.
  • Match each SKU to the features your users actually use, using Setup and adoption reports.
  • Mark features available through more than one SKU, and decide which one you keep.
  • Check whether a new edition already includes Flex Credits or Data 360 capacity you were planning to buy separately.
  • Confirm whether consumption-based agents and per-user add-ons would cover the same users and the same work.

What should I ask my Salesforce account team?

Ask questions that force a written mapping from your current contract to current product names. Verbal assurances about what is included are hard to enforce later.

  • Which of our current SKUs carry Einstein names, and what is each one called now?
  • At renewal, will our Einstein add-ons convert to Agentforce add-ons, or appear alongside them?
  • Which features in our org are in maintenance mode, and what replacement do you recommend?
  • Does our edition already include Einstein Activity Capture, Prompt Builder or AI Models?
  • If we move to a newer edition, which existing add-ons become redundant, and will they be removed?
  • How are Flex Credits and Data 360 credits allocated, and what happens to unused capacity?
  • Can you put the old-to-new mapping in writing as part of the quote?

Product packaging changes often, and terms vary by contract and region. Treat every licensing note in this guide as something to confirm with your Salesforce account team, not as a commitment.

Where should we start if we inherited a mix of both?

Start with an audit of what is switched on and what is being used, then decide on agents. Data, permissions and process usually matter more than which brand is printed on the license.

Our guide to what Agentforce is explains how agents work and what they need. Before a first agent goes live, work through the Agentforce readiness checklist. For an outside review of your org and contract, an Abstrakt health check can compare active Einstein features with what you pay for.

Chris Gooding, President & CEO of Abstrakt Solutions
President & CEO, Abstrakt Solutions
LinkedIn →

Tech Talk

A monthly brief for the people who own Salesforce, AI and revenue technology

What changed in Salesforce and AI this month, and what to do about it.

One email a month. Written by the consultants who deliver the work, not by a marketing team, for the leaders who make the technology decisions.

  • What changed in Salesforce, AI, integration and RevOps, and what it means for your org
  • At least one framework, checklist or reference architecture you can take into a meeting
  • Honest opinions, including when we disagree with what a vendor is selling
  • No sales sequence. We do not sell from this list

Consultant analysis, not vendor recaps. One click to leave.

One email a month. Your industry and your address, nothing else. We never share either, and you can unsubscribe from the bottom of any issue. See what’s in Tech Talk →

Call (314) 916-4095 Book a consultation
Call (314) 916-4095 Book a call