The Architecture of the Agentforce 360 Platform

To understand the licensing, you must first understand the platform. Agentforce 360 is Salesforce's umbrella brand for its AI-native enterprise platform. It integrates five core components: Agentforce (the AI agent runtime that executes tasks autonomously), Data 360 (the data backbone that gives agents contextual grounding), Einstein AI (the generative AI and prediction engine), Salesforce Flows (the low-code automation layer), and Salesforce Foundations (the shared platform services). Together, these components allow organisations to build, deploy, and scale AI agents that interact with customers, employees, and systems with varying degrees of autonomy.

The commercial model for Agentforce 360 is hybrid: some components are per-seat (traditional user licences for Salesforce CRM clouds), some are per-consumption (Data 360 Flex Credits and Agentforce conversations), and some are bundled into edition tiers. The interaction between these billing units creates the complexity that most enterprises underestimate at the point of purchase.

Data 360 Editions and What Each Includes

Data 360 is available in multiple editions, each designed for different levels of data activation complexity. The transition from Data Cloud to Data 360 introduced updated SKUs, and not all legacy Data Cloud features map directly to the new edition structure.

Data 360 Starter Edition

The Starter edition provides basic unified profile management, standard data connectors to Salesforce native sources (Sales Cloud, Service Cloud, Marketing Cloud), and foundational segmentation capabilities. It includes a limited credit allocation for operations outside the free Salesforce-native ingestion pool. Storage is charged separately at approximately $23 per terabyte per month. Starter is appropriate for organisations building initial customer data unification use cases before scaling to more complex activation patterns.

Data 360 Growth Edition

Growth adds advanced segmentation and activation capabilities, real-time profile enrichment, access to Einstein AI model grounding for Agentforce agents, and expanded connector options for third-party data sources. The credit pool allocation is higher than Starter, and growth-tier customers receive access to the Salesforce Digital Wallet — the near-real-time consumption monitoring tool that is the primary mechanism for tracking credit spend before overruns occur.

Data 360 Plus and Enterprise

At the highest tiers, Data 360 includes Data Spaces (logical partitioning for multi-brand and multi-geography data isolation), the zero-copy architecture for direct data sharing with cloud data warehouses such as Snowflake and Databricks without data movement, advanced analytics and insight generation, and enhanced support for Agentforce agent grounding at scale. Enterprise-tier Data 360 is the appropriate choice for organisations running complex, multi-cloud data environments with more than 10 million unified customer profiles or operating across multiple Salesforce orgs.

Agentforce Editions: The Per-Seat and Per-Conversation Mix

Agentforce is available in standalone add-on form and as bundled editions. The edition structure is important because the bundled editions include Flex Credit allocations and Data 360 access that change the total cost economics of a deployment significantly.

Agentforce Add-On ($125 per user per month)

The standalone Agentforce add-on provides access to the Agentforce agent builder, pre-built agent templates, and the ability to deploy agents within the Salesforce environment. It does not include Data 360 access or any Flex Credit allocation — credits must be purchased separately. The add-on is priced for organisations that need basic agent capability without the full Data 360 grounding layer, typically internal-facing automation use cases that do not require deep customer data context.

Agentforce Enterprise ($175 per user per month)

Agentforce Enterprise bundles Sales Cloud or Service Cloud Enterprise features with Agentforce capability and a limited Flex Credit allocation per user. This is the tier that Salesforce positions as the standard enterprise deployment path and the one that most account executives propose as the baseline. The key negotiating point at this tier is the Flex Credit allocation: the bundle includes credits, but the exact allocation and whether it is additive to any existing Data 360 credit pools requires explicit contractual confirmation.

Agentforce Unlimited ($350 per user per month)

Agentforce Unlimited includes all Enterprise features plus expanded Flex Credit allocation, 24/7 Premier Support, unlimited API calls within defined rate limits, and access to advanced Einstein AI capabilities including expanded generative AI model access. The price premium over Enterprise is substantial and requires careful evaluation of whether the additional Flex Credits, support tier, and API limits are genuinely needed.

Agentforce 1 ($550 per user per month)

Agentforce 1 is the flagship edition, designed for organisations building Agentforce as the centre of their entire customer and employee engagement model. It includes one million Flex Credits per organisation per year (not per user — this is a critical distinction), 2.5 million Data Services Credits, full Data 360 access, Einstein AI at full capacity, and unlimited Salesforce Flows automation. The per-org rather than per-user credit allocation means that Agentforce 1 economics are most favourable for high-volume deployment scenarios where the credit pool is shared efficiently across a large user population.

Not sure which Agentforce edition is right for your organisation?

We model the full cost of each edition against your actual use case volumes before you commit.
Request Cost Modelling →

Flex Credits: The Mechanics of Consumption

Flex Credits are the universal consumption currency across Agentforce 360. Understanding what consumes credits — and at what rate — is the prerequisite for any accurate cost modelling.

Free Operations Under Data 360

As of February 2026, the following operations are free of Flex Credit charge: data ingestion from Salesforce Core (Sales Cloud and Service Cloud), Marketing Cloud Engagement, Marketing Cloud Personalization, and Commerce Cloud; alerts and event triggers from Salesforce native objects; and access to the Salesforce Digital Wallet monitoring tool. This free ingestion represents a significant change from the original Data Cloud pricing model and reduces the baseline cost for organisations that primarily work with Salesforce-native data.

Credit-Consuming Operations

Operations that consume Flex Credits include: ingestion from external or non-Salesforce data sources (the credit rate varies by connector type and volume tier); profile unification runs where records from multiple sources are matched, merged, and enriched to create unified customer profiles; segment creation and refresh operations, particularly for large and frequently refreshed segments; Agentforce agent actions — each action (updating a record, calling an external API, generating a summary, executing a Flow) consumes credits at published rates; Einstein AI generative model calls for content generation, summarisation, and classification; and speech and voice services for Agentforce voice agent deployments.

The published consumption rate for Agentforce actions is variable by action type. A simple record update consumes significantly fewer credits than an Einstein-generated case summary or an agent conversation turn that requires multiple data retrieval operations. Salesforce provides a consumption estimator tool, but the estimates are based on simplified assumptions that frequently underestimate actual production consumption by 20 to 40 percent in complex deployments.

Volume Tiering Within a Contract Period

Data 360 credits operate on a monthly volume tiering structure within the contract period: the more credits consumed in a given month, the lower the per-credit rate for the incremental consumption above each tier threshold. This tiering is distinct from the contracted credit pool — it affects how efficiently you consume credits you have already purchased during high-volume periods. Organisations with highly seasonal consumption patterns should model both average and peak monthly volumes to understand their effective credit economics.

Agentforce Per-Conversation Pricing: The Hidden Scaling Risk

In addition to Flex Credits, Agentforce charges separately for conversations — defined as a complete interaction between an AI agent and a user, from the first message to resolution. The list price is $2 per conversation, with volume discounts available at committed annual volumes. This pricing creates a cost structure that is directly proportional to agent adoption success — the better your agents work, the more conversations they handle, and the higher the bill.

The conversation pricing model is most consequential for customer-facing deployments. An Agentforce-powered customer service deployment handling 100,000 conversations per month generates $2.4 million per year in conversation fees alone at list price. Even at a negotiated 50 percent discount, the conversation cost layer adds $1.2 million annually on top of all base licences and Flex Credit consumption.

Agentforce conversations are priced separately from Flex Credits. An organisation running 100,000 AI agent conversations per month pays $2.4M annually at list price — before any base licence or Data 360 costs. Model the conversation layer before deploying at scale.

Conversation Counting and What Constitutes a Conversation

Salesforce defines a conversation as a single session between an agent and a user, regardless of the number of turns (individual messages). A customer who asks an agent three questions in one session and gets three responses generates one conversation charge. A customer who initiates, abandons, and re-initiates generates two charges. Transferred conversations — where an AI agent escalates to a human agent — count as the AI conversation portion only; the human agent interaction falls under the standard Service Cloud licence cost. Salesforce's contract language around conversation counting should be reviewed carefully for edge cases relevant to your specific deployment architecture.

The Annual Uplift and Salesforce Fiscal Year Dynamics

Salesforce's standard Order Form includes an 8 to 10 percent annual uplift clause. For Flex Credit and conversation-based contracts, the uplift applies to the contracted credit pool and conversation allowance, creating compounding cost increases even when actual usage remains flat. A contracted Data 360 credit pool worth $500,000 per year increases to $540,000 in year two, $583,200 in year three, and $629,856 in year four — an accumulated increase of $253,056 over three years without any increase in business usage.

Salesforce's fiscal year ends January 31. The Q3 (October 31) and Q4 (January 31) quarter-ends are the windows of maximum commercial flexibility. The January 31 fiscal year-end is particularly powerful because it represents Salesforce's annual quota attainment deadline. In the five business days preceding January 31, Salesforce account teams receive maximum discount authorisation and have the greatest incentive to close outstanding renewals. Organisations that have done the commercial groundwork — competitive evaluation, consumption modelling, defined negotiating position — and bring their renewal conversation to January quarter-end consistently achieve uplift caps of 4 to 5 percent and per-credit pricing 15 to 25 percent below list.

Client Outcome: In one engagement, a global financial services firm had committed to Agentforce 1 for 800 users without modelling conversation volume. Their projected annual conversation cost at list price exceeded $3.6M. Redress renegotiated the conversation pricing tier, capped annual uplift at 4%, and secured a consumption credit buffer — reducing the three-year cost exposure by $2.1M. The engagement fee was less than 3% of the savings achieved.

Our Salesforce licensing advisory specialists model total cost of ownership — licences, Flex Credits, conversation fees, and uplift — before any commitment is made.

Five Priority Recommendations for Data 360 and Agentforce Buyers

1. Model Consumption Before Committing: Use Salesforce's published credit consumption rates and the Digital Wallet estimator, but add a 30 percent buffer to all projections for first-year deployments. First-year actual consumption consistently exceeds projections as use cases expand beyond initial scope.

2. Separate the Conversation Pricing Negotiation: Flex Credits and Agentforce conversations should be negotiated as separate line items. Volume commitments on one should not automatically determine pricing on the other. Establish dedicated conversation pricing tiers based on your projected deployment scale.

3. Lock Credit Consumption Rates: Require contractual locks on the credit consumption rate per operation type. Salesforce retains the right to modify consumption rates, which effectively reduces the purchasing power of pre-purchased credits without violating the letter of the contract.

4. Negotiate Sandbox Credit Discounts: Data 360 and Agentforce operations in sandbox environments consume credits at 80 percent of the production rate (a 20 percent discount). Ensure this discount is contractually confirmed and that sandbox credits draw from a separate allocation rather than the production pool where possible.

5. Use Salesforce's Fiscal Calendar: Salesforce's most aggressive commercial concessions occur in the final weeks of Q3 (October) and Q4 (January). If your renewal falls in a different quarter, explore whether Salesforce will incentivise an early renewal — with guaranteed pricing continuity — timed to their fiscal quarter-end.

Get the Latest on Data 360 and Agentforce Pricing

Salesforce updates its Data 360 and Agentforce pricing model regularly. Subscribe for quarterly analysis of what's changing and how it affects enterprise contracts.