tools
Snowflake Cortex Analyst
Snowflake Cortex Analyst converts natural-language questions about Snowflake data into SQL and answers for business users.
Cortex Analyst is a managed Snowflake feature for building applications that answer questions about structured data. It uses semantic models or semantic views to provide business definitions, metrics, relationships, and verified examples for SQL generation.
Developers can access it through a REST API or integrate it into applications such as Streamlit apps, Slack, Teams, and custom chat interfaces. Snowflake recommends transitioning to Cortex Agents, which adds unstructured-data retrieval, tool calling, conversation threads, and multi-step orchestration. Generated SQL also incurs standard virtual warehouse charges.
Features
- Converts natural-language questions into SQL
- Uses semantic models and semantic views for business context
- Supports verified queries to improve SQL generation
- Provides a REST API for application integration
- Supports multi-turn conversations
- Runs models inside Snowflake's security perimeter
- Tracks usage through CORTEX_ANALYST_USAGE_HISTORY
- Supports evaluation of semantic views and generated SQL
Use cases
- Build chat applications for self-service business analytics
- Let nontechnical users query structured Snowflake data
- Embed natural-language analytics in internal business tools
- Generate SQL for Streamlit applications
- Evaluate semantic views against verified business queries
Pros
Cons
Pricing
- Starting price
- 67 Platform Credits per 1,000 messages
- Pricing checked
- 2026-09-19
Cortex Analyst API
67 Platform Credits per 1,000 messages
- Standalone Analyst API billing
- Generated SQL execution incurs standard virtual warehouse charges
Cortex Analyst via Cortex Agents
AI Credits per one million Tokens
- Token-based pricing
- Rate depends on the selected model
- Costs can be additive with other agent tools