Automation Cloud admin guide

Last updated Nov 7, 2025

Context Grounding licensing

This page provides details on how Context Grounding requests are quantified through licensing units.

Unified Pricing

Note: The licensing information in this section applies to you if you are on Unified Pricing. For details, refer to Licensing plan framework.

Context Grounding consumes Platform Units for each request (search, summarization, or RAG operations) as it is executed through GenAI Activities. The consumption of Platform Units occurs in the following way:

Basic search and retrieval is free. This includes any time the index is accessed by a GenAI activity or an agent.
Basic ingestion is free. Applies to digital text input only (excluding images and tables). Limited to 1.5 million tokens input tokens in a single index (approximately 1,000 pages in a single index).
Advanced ingestion is charged at 0.4 Platform Units per 15,000 tokens of input (or 10 pages). Applies to multi-modal documents such as scans, tables, or images; large documents typically containing around 1.5 million tokens or more than 1,000 pages; or extensive unstructured text, like JSON files exceeding 6 million characters.
DeepRAG queries are charged at 0.2 Platform Units per 30,000 tokens of input, equivalent to ~20 pages. This enables synthesized insights across documents.

Note: Context Grounding requests made from Autopilot or Agents consume licensing units based on the host product:

In Autopilot for Everyone, each action—whether or not it includes a Context Grounding query—consumes 1 Autopilot action.
In Agents, executions consume Platform Units.

Note: For DeepRAG, Platform Unit consumption may scale with input size (e.g., per 30K tokens or ~20 pages), though detailed per-unit pricing may vary depending on the surface.

Flex Licensing

Note: The licensing information in this section applies to you if you are on the Flex plan. For details, refer to Flex licensing plan framework.

Context Grounding consumes AI units for each request (search, summarization, or RAG operations) as it is executed through GenAI Activities. The consumption of AI units occurs in the following way:

Basic search and retrieval is free. This includes any time the index is accessed by a GenAI activity or an agent.
Basic ingestion is free. Applies to digital text input only (excluding images and tables).
Advanced ingestion is charged at 2 AI Units per 15,000 tokens of input (or 10 pages). Applies to multi-modal documents such as scans, tables, or images; large documents typically containing around 1.5 million tokens or more than 1,000 pages; or extensive unstructured text, like JSON files exceeding 6 million characters.
DeepRAG queries are charged at 1 AI Unit per 30,000 tokens of input, equivalent to ~20 pages. This enables synthesized insights across documents.

Note: Context Grounding requests made from Autopilot or Agents consume licensing units based on the host product:

In Autopilot for Everyone, each activity—whether or not it includes a Context Grounding query—consumes 1 Autopilot action.
In Agents, executions consume AI units.

Entitlements

Community users are subject to the following quotas:

Index allocation: 10 indexes per tenant.
Data ingestion: Daily limits based on the number of ingested pages.
Query allowance: Up to 100 free queries per day, after which consumption starts.

Trial and Enterprise tenants receive:

Index allocation: 10 indexes per tenant.
Query allowance: Up to 100 free queries per day, after which consumption starts.
DeepRAG: 2 free queries per day (equivalent to 50 basic searches per call), after which consumption starts.

Note: Advanced ingestion is not available for Community and Trial accounts.

On this page