Usage and Billing for Agentforce Vibes Considerations

To use Agentforce Vibes, consider the supported functionality, usage, limitations, allowances, limits, and other issues.

Both billing paths require Einstein to be enabled. See Set Up Entitlement for Agentforce Vibes. From Setup, in the Quick Find box, enter Einstein Setup, and then select Einstein Setup. Make sure that Einstein is enabled. If it isn’t, select Turn on Einstein.

Agentforce Vibes currently supports these models and adds support for new models on an ongoing basis.. For more information about model types, see Large Language Model Support.

Usage TypeModels AvailableNotes
UnmeteredGPT-5 (Included), Claude Sonnet 4.5 (Premium), Claude Sonnet 4.6 (Premium)Limit for premium models is 2,500 requests or 36 million tokens, whichever is reached first. The Included model is subject to contractual fair use. See Unmetered Platform Developer and Admin AI User Permission Set License for more information.
Flex Credit (Metered)GPT-5.4, Claude Opus 4.5, Claude Opus 4.6Opus uses Advanced prompts

Token usage is tracked independently of model selection. Switching between models does not reset or bypass usage limits. Flex Credit multiplier depends on the Prompt usage type defined for the selected model. See Large Language Model Support for models and associated Usage Types.

Users without the Unmetered SKU can access all supported models via Flex Credits.

Agentforce Vibes uses generative AI and Data Cloud to ingest, store, and process data. You’re billed in these categories when you use Agentforce Vibes. Before deployment, work with your Salesforce account team to confirm license availability and plan credit usage.

This feature has access to Digital Wallet, a free account management tool that offers near real-time consumption data for enabled products across your active contracts. Access Digital Wallet and start tracking your org’s usage. To learn more, see About Digital Wallet.

Metered usage of Agentforce Vibes is billed via Flex Credits. To assign the required licenses and permission sets to users, see Set Up Entitlement for Agentforce Vibes.

CARDTypeDESCRIPTION
Flex CreditsPrompt UsageUsage is calculated using Prompt Usage Types. The Prompt Usage Type (Starter, Basic, Standard, or Advanced) is determined by the large language model (LLM) used. Each Prompt Usage Type includes the processing of up to 2,000 input and output tokens. Usage is counted in 2,000-token increments and rounded up. Prompts that exceed this limit are metered as multiple Prompt Usage Types, with each additional 2,000-token increment counted separately. For example, a prompt with a total of 6,500 input and output tokens is metered as four units of the applicable Prompt Usage Type. The Flex Credit multiplier applied to each Prompt Usage Type is defined in the current Rate Card. Tokens are units of data processed by AI models.

For more information about Flex Credits, see Flex Credits Billable Usage Types.

Usage of Agentforce Vibes is available on an unmetered basis. Unmetered usage doesn’t consume credits. To assign the required licenses and permission sets to users, see Set Up Entitlement for Agentforce Vibes.

For more information about these permissions, see Unmetered Agentforce for Platform AppDev Permission Set License. For information about assigning these unmetered permissions to users, see Configure Unmetered User Based AI Permissions.

From Summer ’26 release onwards, you can enable unmetered access to Agentforce Vibes in scratch orgs using the AgentforceVibesUnmeteredAccess feature in your scratch org definition file.

Prerequisites

Your Dev Hub org must be enabled for unmetered Agentforce Vibes usage.

Example scratch-def.json

Scratch orgs created with this feature have unmetered access limits as documented here.

Developer Edition orgs provide limited free access to Agentforce Vibes. The limits for Agentforce Vibes in DE orgs is 110 requests or 1.5M tokens per org, whichever you hit first. This limit applies for one month, after first time use, and it doesn’t refresh.

Agentforce Vibes Pricing Large Language Model Support