Salesforce Flex Credit Estimator

Estimate your Salesforce AI consumption costs for LLM prompts, Agentforce conversations, and vector database chunking.

Prompt Cost Estimator

Estimate flex credits and cost for individual LLM prompt calls based on token count and model tier.

Inputs

Model Type

View credit tiers | Model support

Total Tokens

~1,500 words

Prompt Generations per Month

Expected monthly volume

Results

Flex Credits per Prompt Generation

Cost per Prompt Generation (USD)

$0.0200

Monthly Totals

4 credits

$0.02

Calculation:

Token Multiplier = ceil(2,000 / 2,000) = 1

Credits per Generation = 4 × 1 = 4

Monthly Total = 4 × 1 = 4

Note: Each 2,000-token chunk is metered once. Costs scale with your selected tier and total token count.

Conversation Cost Estimator (Agentforce)

Estimate monthly flex credits and costs for ongoing Agentforce agent conversations. Each agent action consumes 20 credits and costs $0.10.

Inputs

Number of Users

How many users or agents will be having conversations

Conversations per User / Month

Average number of conversations each user has per month

Avg. Agent Actions per Conversation

Average actions the agent takes in each conversation (searches, updates, API calls, etc.)

Results

Total Flex Credits / Month

100,000

Estimated Monthly Cost (USD)

$500.00

Cost per Conversation (USD)

$0.50

Calculation:

Total Actions = 10 users × 100 conversations × 5 actions

Total Actions = 5,000

Total Credits = 5,000 × 20 = 100,000

Monthly Cost = 5,000 actions × $0.10 = $500.00

Standard Action Pricing:

Each standard action = 20 flex credits = $0.10 USD
Each bundle costs $500 and provides 100,000 Flex Credits

Token Consideration:

Each standard action includes processing of up to 10,000 tokens
Actions exceeding this limit count as separate actions for each 10,000 tokens
Example: 20,001 tokens = 3 standard actions
Actions with lengthy prompts may count as multiple actions if the 10,000-token limit is exceeded

Vector DB Chunk Estimator

Estimate the number of chunks generated when preparing text for a vector database. Experiment with different chunking strategies to optimize storage and retrieval costs.

Input & Configuration

Character Count

Or paste text below to count automatically

Paste Text Here

No text pasted

Chunking Parameters

Chunk Size (units)

Overlap (units)

Prepended Context (characters)

Metadata or context added to each chunk (e.g., document title, section header)

Results

Estimated Units

Based on ~4 chars/unit heuristic (0 characters)

Your Custom Chunk Size (1024 units)

1 chunks

Preset Chunk Sizes

512 units:1 chunks

768 units:1 chunks

1024 units:1 chunks

How it's calculated?

Unit Estimation: We convert characters to "units" (an abstraction for tokens) using the heuristic: units = ceil(characters / 4)

Chunking Formula:

Calculate prepended units: prependUnits = ceil(prependChars / 4)
First chunk content: contentInFirstChunk = chunkSize - prependUnits
If totalUnits ≤ contentInFirstChunk, return 1 chunk
Subsequent chunk content: contentInSubsequentChunks = chunkSize - prependUnits - overlap
Remaining units: remainingUnits = totalUnits - contentInFirstChunk
Additional chunks: additionalChunks = ceil(remainingUnits / contentInSubsequentChunks)
Total: 1 + additionalChunks

Character-to-Unit Examples:

Characters	Estimated Units
1,000	250
5,000	1,250
10,000	2,500

Disclaimer: This tool provides estimates only. Actual consumption may vary based on your specific implementation, model selection, and usage patterns. For official pricing and detailed information, please visit the Salesforce Agentforce Pricing Page and the Einstein Credit Documentation.