
Snowflake CoCo
Your data-native AI coding agent. Turn your complex data engineering, analytics and AI workflows into simple conversations.
Get started in your browser
Overview
Build with Snowflake CoCo wherever you work

Desktop
Snowflake CoCo Desktop
A Snowflake-native desktop IDE that offers one governed surface to build across your data stack.

Terminal
Snowflake CoCo CLI
A terminal-native AI coding agent that connects your local development environment to your data stack.
Browser
Snowflake CoCo in Snowsight
A persistent coding agent available across Snowflake Workspaces, Notebooks and other Snowsight UI workflows.

Existing Tools
Snowflake CoCo Extensions
Run CoCo directly in your IDE, bringing data-native agentic coding into tools you already use.
Benefits
Accelerate Your Enterprise Data Lifecycle with Snowflake CoCo
Build exponentially faster from a simple conversation
Create data pipelines, analytics, ML models, apps and AI agents in plain language.
Do reliable work with an agent that knows your stack
CoCo can read your catalog, lineage and RBAC policies, so generated code references real objects with correct permissions.
Fit into your tools and workflows without imposing new ones
Work wherever you already work — Snowflake, your terminal, or IDE — with a governed agent for enterprise scale.
Faster innovation
Imagine, build and ship at the speed of your ideas
- Identify data and build pipelines across dbt, Apache Airflow,®1 Postgres, Spark and AWS Glue in natural conversations.
- Generate fully executable ML pipelines autonomously, ready to run directly in Snowflake Notebooks.
- Create apps and agents grounded in your enterprise data, governance and compute context.
"Snowflake CoCo is becoming a core part of how we operate as a company. We’re rolling it out across the entire organization, not just our data team. Because it understands our data, our environment, and our governance, teams can build and automate workflows on top of trusted data without needing specialized expertise."
Matt Luizzi

Strong data foundation
Tap into an agent harness designed for data, not just code
- CoCo’s behavior is grounded from the first prompt because it’s familiar with your enterprise context including catalog, lineage, RBAC, compute and pipeline dependencies.
- Ditch generic code wrappers for built-in, Snowflake-specialized tooling like semantic catalog search, data diffing, sandboxed runtimes and more.
- Simplify complex multi-step data tasks automatically. Orchestration coordinates sub-agents and routes between models to balance cost and accuracy automatically.
Native and embedded
CoCo is available where you build
- Work natively in Desktop, CLI or Snowsight, with full enterprise data context across different surfaces.
- Integrate into your existing IDE or tools through the CoCo extension for VS Code, CoCo plugin for Claude Code or Agent Client Protocol, supporting 30+ editors.
- Embed CoCo into agents and automated workflows via the SDK, MCP server and Async API.

“CoCo helps us reduce friction in everyday data and AI development while maintaining the controls and oversight we need in a regulated environment. Our teams can build faster with the context they need to be successful.”
Vibhor Gupta


Enterprise-ready by design
Leverage built-in extensibility, interoperability and customization
- Manage permissions and usage policies through centralized configuration and fine-grained controls.
- Connect to Jira, GitHub and other agents in your developer toolchain via out-of-the-box Model Context Protocol (MCP) support. Embed into agents and automated workflows via the SDK, MCP server and Async API.
- Build, share and refine specialized agent skills, or migrate existing workflows from other coding agents using the open agents.md framework.
- Choose between models, including Claude Opus 4.7, Claude Sonnet 4.7 and GPT 5.4, for quality, latency and cost.

“CoCo helps our engineers improve the performance of our business intelligence tools, meaningfully reducing the time it takes to improve quality and speed of natural language query responses.”
Tony Leopold
Ready to get started with Snowflake CoCo?
Resources
Explore CoCo Developer Guides
Learn how to build using Snowflake's native AI capabilities, with CoCo as your AI coding assistant.
Get Started
Take the next step with Snowflake
Start your introductory trial with $40 in free credits²
- Accelerate end-to-end development
- Context-aware by default
- Open, extensible and production-ready
Snowflake CoCo
Frequently Asked Questions
Snowflake CoCo is the first truly Data-native AI coding agent, built to speed time to value across modern data stack. Deeply aware of your enterprise's data, compute, governance, and operational semantics. It works directly inside Snowflake, using your governed data, metadata and permissions to generate accurate, production-ready code so you can move from idea to execution faster.
Yes. Snowflake CoCo is simply the new name for Cortex Code. The product functionality and architecture remain the same. The rename reflects a broader vision for AI-powered development within Snowflake, but existing capabilities and workflows are unchanged.
Together with Snowflake CoWork, CoCo provides customers with an agentic control plane that lets them manage and orchestrate agent behavior.
Snowflake CoCo is billed based on token consumption. Snowflake also offers trial access with free credits for new users. For the latest pricing details: View Snowflake pricing for the latest details.
CoCo is available across multiple interfaces so you can work where you already build:
- Desktop App: Full agentic development on data available in a native desktop IDE
- Snowsight: Built into Snowflake for SQL, notebooks and analytics workflows
- CLI: Runs in your terminal for local development, automation and CI pipelines
All interfaces share the same core capability: CoCo understands your Snowflake environment and generates code grounded in your real data and permissions.
CoCo Skills are built-in specialized workflows that allow CoCo to complete complex data and development tasks for you. These include things like:
- Searching your data catalog
- Generating and running SQL
- Parsing dbt projects
- Comparing datasets and validating outputs
You do not need to (although you can) invoke Skills manually. Simply describe what you want in natural language and CoCo automatically selects and coordinates the right tools. To learn more, explore Snowflake CoCo documentation.





