Salesforce is redesigning the commercial structure around its flagship enterprise applications, introducing a three-tier lineup that embeds artificial intelligence more deeply into the software licenses customers purchase for sales, customer service and industry-specific operations.

The company announced Core, Advanced and Max editions on September 3 for Agentforce Sales, Agentforce Service and Agentforce Industries. Rather than requiring enterprises to assemble separate licenses for AI agents, collaboration, analytics, security and support, Salesforce is packaging more of those components together under a single edition structure. The company is positioning the move as a simplification of a product catalog that has expanded rapidly as Salesforce added Agentforce, Data Cloud capabilities, Slack integrations, Tableau-based analytics and other AI-oriented services.

The most visible change is the pricing ladder. Salesforce listed Core Editions at $195 per user per month, Advanced Editions at $395 per user per month and Max Editions at $550 per user per month. Salesforce said Core offers about 70% more value than its legacy Enterprise Edition, while Advanced offers more than 50% more value than the legacy Unlimited Edition. The Max price is unchanged from the $550 list price attached to Agentforce 1, but Salesforce says the new package delivers nearly 60% more value.

The company’s value percentages represent Salesforce’s assessment of the additional products and capabilities incorporated into each bundle, rather than a reduction in the cash subscription price paid for the lower tiers. Industry coverage following the announcement noted that the $195 Core level compares with a $175 legacy Enterprise list price, while Advanced at $395 compares with approximately $350 for the previous Unlimited tier. That makes the packaging change both an expansion in functionality and, for customers buying comparable lower-tier licenses at current list prices, a higher entry price.

Salesforce said existing customers on legacy editions will retain their current pricing rather than being automatically moved to the newly announced rates. That distinction is important for large companies that negotiate multiyear Salesforce agreements and often maintain different editions across business units. The announcement therefore has more immediate implications for new purchases, upgrades and future contract negotiations than for customers whose existing agreements remain in force.

At the top of the lineup, Salesforce is making a different calculation. Max keeps the $550-per-user monthly price associated with Agentforce 1 while increasing the quantity of bundled services. Existing Agentforce 1 Edition customers can move to the new Max edition without an additional subscription-price increase, according to Salesforce. That provides the company with a migration path for customers that previously committed to its highest-end AI package while giving them more reasons to consolidate additional workloads inside the Salesforce ecosystem.

A central feature of the new structure is the inclusion of Flex Credits, Salesforce’s unit for metering certain Agentforce activities. Core Editions include 500,000 Flex Credits, Advanced Editions include 1 million and Max Editions include 2.75 million. The allocations give customers an initial pool of capacity with which to build, test and operate AI agents before purchasing additional consumption.

Flex Credits have become one of Salesforce’s principal mechanisms for pricing agentic computing. Under the company’s broader Agentforce pricing structure, customers can buy 100,000 Flex Credits for $500, while a standard Agentforce action typically consumes 20 credits. Salesforce also offers conversation-based pricing and other purchasing arrangements, producing a hybrid commercial model in which traditional per-user subscriptions coexist with metered AI activity.

That approach reflects a wider challenge confronting enterprise software companies. Conventional cloud applications have historically been priced around users, seats, modules or transaction volumes. Autonomous AI systems consume computing resources differently. A single employee may initiate a workflow that causes an agent to execute numerous actions across databases, applications and external systems. Vendors therefore need a way to charge for the expanding machine workload without making enterprise budgeting too difficult to forecast.

Salesforce’s new editions attempt to bridge those models by including an initial quantity of AI consumption inside a recognizable per-user subscription. For customers, that can lower the procedural barrier to experimentation because AI capabilities and some consumption capacity arrive with the broader platform rather than requiring a separate procurement exercise. For Salesforce, the structure can create additional consumption revenue if successful Agentforce deployments exceed the credits bundled with an edition.

The new packaging extends beyond Agentforce itself. Salesforce said every edition tier includes Slack and Slackbot, embedded agentic analytics, data-security capabilities and Premier Success Plans. Tableau Next functionality is also part of the broader bundle, linking data analysis more closely to AI agents that can interpret information and initiate actions based on enterprise data.

The Slack inclusion is strategically significant because Salesforce increasingly describes conversational interfaces as a primary environment in which employees will interact with both human colleagues and software agents. Rather than forcing users to move between a CRM application and a separate AI interface, Salesforce wants agents to work within Slack conversations and operational workflows. That can also make the company’s $27.7 billion acquisition of Slack more central to its AI monetization strategy.

Salesforce is similarly integrating analytics into the agentic architecture. The company’s concept of embedded agentic analytics combines conversational AI with semantic models and preconfigured analytical applications. The objective is to allow agents not only to retrieve information but also to interpret business context and initiate actions. For enterprise buyers, Salesforce is effectively arguing that AI agents become more valuable when they operate across CRM records, analytics, collaboration tools and governed corporate data inside one platform.

Security and governance are another prominent part of the packaging. Salesforce said the editions include controls intended to protect sensitive data, manage permission changes and support broader enterprise security requirements. The emphasis reflects a major obstacle to deploying autonomous software inside large organizations: agents may need permission to access customer records, execute workflows or modify business data, creating risks that are more complex than those associated with a conventional chatbot that only generates text.

The specific capabilities also vary by application and edition. In Agentforce Sales, Salesforce said sales-oriented agents can help prospect for customers, qualify leads, engage buyers and move opportunities through the pipeline. Momentum, a capability designed to capture activity and keep CRM records updated, is included across Core, Advanced and Max. Sales Programs, which provides structured sales playbooks and incentive-related capabilities, begins with the Advanced level and is also included in Max.

For customer-service operations, Salesforce is using the edition structure to differentiate increasingly autonomous workflows. Agentforce Help Agent is included in Advanced and Max and uses an outcome-based pricing approach under which customers are charged when issues are resolved. Max adds higher-end capabilities including Agentforce Service Rep Assistant, Workforce Engagement and Agentforce IT Service, giving Salesforce a way to reserve more sophisticated employee-agent and operational-management functions for its premium tier.

The service strategy illustrates how Salesforce is experimenting with several AI pricing units at once. A company may pay a conventional monthly amount for an employee license, consume Flex Credits when agents perform defined actions and encounter outcome-based charges for certain autonomous service functions. The result offers flexibility but also increases the importance of workload modeling when chief information officers compare the total cost of AI deployments.

For software buyers, the headline edition price therefore represents only part of the potential economics. Organizations evaluating Agentforce will need to consider how many employees require licenses, which functionality belongs in each edition, how rapidly bundled Flex Credits could be consumed, and whether high-volume use cases generate additional consumption charges. Enterprises that move from pilots to autonomous agents handling significant portions of sales administration, customer support or internal IT could see consumption become a more material component of Salesforce spending.

That is one reason the larger Flex Credit allocations matter. A pool of credits included with the subscription gives organizations room to experiment without treating every initial agent interaction as a separate budget decision. Salesforce can then use measured consumption to demonstrate the operational value of agents and encourage customers to expand deployments. The strategy resembles the broader cloud-computing model in which committed software relationships create a base of recurring revenue while usage generates incremental spending.

Agentforce Industries adds a vertical dimension to the same strategy. Salesforce said the industry editions combine sales and service functionality with specialized data models, governance controls, preconfigured workflows and agents designed around vertical context. The company cited sectors including financial services, healthcare and the public sector, where regulatory requirements and specialized business processes can make general-purpose AI deployments more difficult.

Salesforce said new pricing for Agentforce Industry Editions will take effect later in the fall. That creates a phased commercial transition and gives customers in regulated sectors additional time to assess how the packaging affects their existing Salesforce environments. The company also cautioned that pricing, packaging and regional availability can vary and are governed by customer agreements.

The September restructuring builds on pricing changes Salesforce introduced during the earlier stages of the Agentforce rollout. In 2025, the company expanded beyond a simple conversational charging model by introducing Flex Credits, per-user Agentforce licenses and Agentforce 1 Editions. Flex Credits were intended to align billing with actions performed by agents, while higher-end licenses offered more predictable access for employee-facing AI use cases. The new Core, Advanced and Max structure moves another step toward incorporating those capabilities directly into Salesforce’s mainstream commercial architecture.

That evolution is significant for Salesforce’s competitive position. Microsoft, ServiceNow, Oracle, SAP and other enterprise software providers are also embedding generative and agentic AI throughout established application suites. As technical differences narrow, packaging can become a major competitive lever. Vendors that make AI easier to purchase, govern and integrate with existing workflows may be able to accelerate adoption even when customers remain cautious about large-scale autonomous deployments.

For Salesforce, bundling also increases the strategic connection among products that were historically bought or managed separately. Slack becomes an operating interface for agents, Tableau contributes analytics, Salesforce’s data layer provides context, security products govern access and Premier Success supports implementation. The company is effectively selling those components as an integrated agentic stack rather than a collection of adjacent applications.

The commercial risk is that customers may view the broader bundles as paying for functionality they are not ready to use. Enterprises remain at very different stages of AI adoption, and many are still running controlled pilots rather than production deployments involving large numbers of autonomous actions. Organizations with limited Agentforce usage may focus more on the higher Core and Advanced list prices than on the theoretical value of additional bundled products.

Conversely, companies already using Slack, analytics, Salesforce AI and premium support may find the consolidated editions easier to justify because separate software components are being brought together in one contract structure. The economics will vary substantially depending on existing discounts, current add-ons, user counts and AI consumption patterns, making negotiated enterprise agreements more consequential than published list prices alone suggest.

The Max tier provides the clearest indication of where Salesforce wants sophisticated customers to move. Keeping the headline price at $550 while expanding AI credits and bundling additional functionality creates an incentive for Agentforce 1 customers to migrate rather than reassess the entire software stack. Salesforce has also said future Max editions are expected to include an allocation for Headless 360 capacity, which would extend Salesforce services into more agent-driven applications and experiences.

The broader objective is to make autonomous agents a standard component of enterprise software purchasing. Instead of presenting Agentforce principally as an optional AI product that sits beside Salesforce’s established applications, the Core, Advanced and Max structure embeds AI, analytics, collaboration and governance into the definitions of the editions themselves.

That transition could have long-term implications for Salesforce’s revenue mix. Seat-based subscription revenue remains the foundation of the business, but growing Agentforce activity could add a consumption component tied to the amount of work performed by software agents. If customers increase their dependence on autonomous workflows, Salesforce could capture both recurring license revenue and incremental AI usage without requiring customers to adopt an entirely separate platform.

The immediate impact, however, will be felt in procurement departments and CIO budget discussions. Customers approaching renewals or planning new deployments will need to compare the higher functionality of Core and Advanced with their legacy configurations, calculate likely Flex Credit demand and determine whether the $550 Max edition offers enough bundled capability to justify moving more users to the premium level.

Salesforce’s September 3 announcement therefore represents more than a product-name change. It is a further attempt to redefine how enterprise AI is packaged and paid for, using conventional software editions as the foundation while introducing AI consumption inside them. Whether the model proves simpler for customers will depend on actual Agentforce usage, but the direction is clear: Salesforce is moving agentic AI from the edge of its pricing architecture toward the center of its enterprise software business.