Agentforce Review: Salesforce's Agentic AI Platform, Pricing & Features
- Agentforce is Salesforce's agentic AI platform for building and deploying autonomous agents that act on your Salesforce data, the market's most commercially proven GaaS product, reportedly at ~$800M ARR and ~29,000 customers by mid-2026.
- Two pricing models: $2 per conversation for customer-facing agents, or Flex Credits ($500 per 100,000 credits; ~$0.10 per action) for usage-based billing across employee, customer, and voice agents.
- Best for organizations already standardized on Salesforce. Weakest as a vendor-neutral layer across many non-Salesforce systems.
What Agentforce is
Agentforce is Salesforce's entry into Agentic AI as a Service (GaaS): instead of a human working inside a CRM dashboard, an Agentforce agent receives a goal, reasons over your Salesforce data, calls tools and workflows, and completes the task, resolving a support case, qualifying a lead, updating records, with a human in the loop only where you set guardrails.
It sits directly on the Salesforce platform, which is the whole point of it. Agents are grounded in your existing CRM data through Data Cloud, built in a low-code Agent Builder, and governed by the same permissions and audit trail as the rest of your Salesforce org. That tight coupling is Agentforce's biggest strength and its main limitation, and it's the lens for everything below.
What Agentforce does well
Agents are grounded in real CRM data out of the box, so they answer from your accounts, cases, and knowledge rather than a generic model. The low-code Agent Builder and Prompt Builder let non-engineers assemble and test agents, and because everything runs inside Salesforce, security review, permissions, and logging reuse controls your team already trusts. For customer service, the most mature agentic use case, Agentforce has the deepest reference base of any single vendor.
Agentforce pricing
Salesforce revamped Agentforce pricing around Flex Credits in 2025, and in 2026 there are effectively two ways to pay, plus per-user licensing for employee agents.
| Model | What you pay | Best when |
|---|---|---|
| Conversations | $2 per conversation (fixed) for customer-facing agents | Volume is predictable and you want a simple per-outcome number |
| Flex Credits | $500 per 100,000 credits; standard action = 20 credits (~$0.10), voice action = 30 credits (~$0.15) | Usage varies, or you're mixing employee + customer + voice agents |
| Agentforce User License | $5 / user / month (consumes Flex Credits) | Rolling agents out to employees |
| Add-ons | ~$125-$150 / user / month, unlimited usage for licensed employees | Heavy internal usage where metering is a hassle |
| Agentforce 1 Editions | from ~$550 / user / month, bundled Flex Credits + Data 360 credits | Enterprise-wide standardization |
There's also a free on-ramp: Salesforce Foundations gives Enterprise Edition (and above) customers 200,000 Flex Credits, 250,000 Data Cloud credits, and access to Agent Builder and Prompt Builder at no extra cost, enough to prototype before committing budget.
The honest read on pricing: the "$2 per conversation" headline is clean, but real costs are driven by how many actions an agent takes per conversation and whether you're on Flex Credits underneath. Model your expected actions-per-task before signing, that, not the sticker price, determines the bill. Always confirm current numbers on Salesforce's pricing page, as this space is repricing fast.
Best for / not for
Best for: organizations already standardized on Salesforce that want agents acting on existing CRM data, with governance and security handled the Salesforce way. Customer service and sales-ops teams are the clearest fit.
Not for: teams that want a vendor-neutral orchestration layer coordinating agents across many non-Salesforce systems, or those uncomfortable with deeper platform lock-in. If your system of record isn't Salesforce, Agentforce's core advantage, native data grounding, largely disappears.
How Agentforce compares
Agentforce's nearest rival by distribution is Microsoft's Copilot Studio, which trades Salesforce's CRM-native grounding for reach across the Microsoft 365 estate. For a full breakdown see our Agentforce vs Microsoft Copilot Studio comparison, and for the wider field, the best agentic AI platforms of 2026.
Frequently asked questions
How much does Agentforce cost? Customer-facing agents are listed at $2 per conversation; alternatively, Flex Credits cost $500 per 100,000 credits (~$0.10 per standard action). Employee agents start at a $5/user/month license, with bundled Agentforce 1 Editions from ~$550/user/month.
Is there a free version of Agentforce? Salesforce Foundations includes 200,000 free Flex Credits and access to Agent Builder for Enterprise Edition customers and above, enough to prototype.
Is Agentforce an example of GaaS? Yes. It packages autonomous agents as a priced product that completes work, which is the defining shape of Agentic AI as a Service (GaaS).
Do I need Salesforce to use Agentforce? Effectively yes. Its value comes from grounding agents in Salesforce/Data Cloud data; without Salesforce as your system of record, most of its advantage is lost.
Ready to evaluate Agentforce? Start with the free Salesforce Foundations tier to prototype, then model your actions-per-conversation before committing. Compare it against the full field →
Sources - Salesforce, Agentforce Pricing, official pricing (Flex Credits, $2/conversation) - Salesforce Newsroom, Flexible Agentforce Pricing, Flex Credits launch - Constellation Research, Salesforce revamps Agentforce pricing, analysis - Domo, Best Agentic AI Platforms 2026, reported ~$800M ARR / ~29,000 customers figure Last fact-checked: Jul 4, 2026 by Andrew Jamerson