PromptQL Pricing and Billing Documentation
Last Updated on 27th February, 2025
On this page, we provide detailed explanations regarding the pricing and billing concepts related to PromptQL.
PromptQL User
The definition of a PromptQL User for billing purposes is dependent on how the customer implements access to PromptQL.
1. PromptQL Playground Access
When PromptQL is accessed via the PromptQL Playground, a PromptQL User is defined as any authenticated Hasura DDN user that accesses the PromptQL Playground, either through the Hasura DDN Console or an embedded element in a custom user interface.
2. PromptQL API Access
When PromptQL is accessed via the PromptQL API, a PromptQL User is defined as a unique entitlement to the underlying supergraph (data) powering PromptQL, as implemented by the customer.
Example Scenario:
Shared entitlement across multiple users:
If an AI agent using the PromptQL API serves 1,000 users, but all users share the same entitlement to the underlying data, this counts as one PromptQL User.Entitlement at the organization level:
If 1,000 users are served across 20 different organizations, and entitlements are assigned at the organization level, this counts as 20 PromptQL Users.Unique entitlements per user:
If each user has a unique entitlement, based on individual data, this would count as 1,000 PromptQL Users.
3. Combined Playground + API Use
In the case of customers using both PromptQL Playground and PromptQL API, the total number of PromptQL Users for billing purposes is the sum of users from both access methods.
"PromptQL Playground" refers to a User Interface component within the Hasura DDN Console that enables users to query and interact with data connected to a supergraph through natural language inputs.
"PromptQL Playground" refers to a User Interface component within the Hasura DDN Console that enables users to query and interact with data connected to a supergraph through natural language inputs.
PromptQL Consumption
When using PromptQL through either the PromptQL Playground or the PromptQL API, there are two primary components that contribute to the consumption cost:
PromptQL Programs
Each interaction with PromptQL may involve the creation and execution of one or more PromptQL programs. The total number of programs executed can vary depending on the complexity of the task, with more complex tasks requiring the execution of multiple programs.LLM Tokens
If you are using the Hasura-provided LLM for generating and executing the query plan, the LLM cost is calculated based on input and output token usage. The number of tokens used depends on the complexity of the task.Note: The LLM cost does not apply if the customer is using their own LLM instead of the Hasura-provided LLM.
Example Consumption Scenarios
Simple Data Interaction:
A simple data interaction may consume 2,000 tokens and result in the execution of 1 program.Complex Data Interaction:
A complex task involving extensive data and multiple exchanges may execute 10 programs and consume 10,000 tokens.
