56 releases · 4 this month
Snowflake added a background boolean parameter to CORTEX_AGENT_RUN and CODING_AGENT tools in its managed MCP server to enable non-blocking, pollable agent executions.
Inference, observability, and cost management for Cortex AI Gateway are now generally available. The gateway gives AI clients in your organization one governed endpoint to reach models, records a trace of every request, and attributes spend to the users who made it.
Level of detail (LOD) expressions are now in preview for semantic views. An LOD expression calculates a value at a grain that's different from the query grain. You can use LOD expressions to:
Cortex Training is now available in preview, allowing users to customize open-weight LLMs using a serverless service within Snowflake.
The following OpenAI models are now in public preview in AI Functions:
Snowflake-managed MCP servers can now expose a coding agent that provides a managed coding sandbox attached to a conversation thread without requiring a stored agent object.
Starting October 20, 2026, Cortex Fine-tuning will no longer support the creation of new fine-tuned models in AWS US East 1 (N. Virginia).
We are pleased to announce the preview of Tools for Cortex AI Gateway, which connects AI clients to the business tools your organization already uses through the Model Context Protocol (MCP). Supported clients include CoCo Desktop, CoCo CLI, Claude, and Cursor.
Cortex AI Gateway is now available to accounts in Microsoft Azure and Google Cloud Platform (GCP) commercial regions, and in the Amazon Web Services (AWS) Asia Pacific (New Zealand) region.
The gemini-4-argon model is now available in Public Preview in AI Functions with AI_COMPLETE.
The claude-haiku-5-5 model is now available in Public Preview in the Cortex REST API.
Data lineage now includes the following relationships:
The Cortex Agent code execution tool is now generally available. The code execution tool is a built-in tool that enables an agent to execute code during a conversation, so your agents can run scripts to process data, perform calculations, and produce visualizations.
Observability for Cortex Agents in a Snowflake Native App now includes the following capabilities:
With this release, DEFINE SEMANTIC VIEW in DCM Projects is generally available. You can declaratively manage semantic view definitions, including tables, relationships, facts, dimensions, metrics, AI instructions, and verified queries.
Snowflake Cortex AI Functions now automatically caches and reuses the results of AI_COMPLETE calls. When a query makes an identical AI_COMPLETE call more than once, Snowflake can return the stored result instead of running inference again, which lowers cost and speeds up query execution.
The Cortex Agents Compact API is now available in preview. Use the agent:compact endpoint to summarize a conversation and pass the compact representation to subsequent agent:run requests. Compacting conversation history reduces token consumption and helps conversations stay within the model context window.
The following Cortex Agents object enhancements are now generally available:
We are pleased to announce the preview of Cortex AI Gateway, the control plane where you govern how AI applications and third-party agents in your organization reach models.
Snowflake Cortex AI_SUMMARIZE is now available in public preview, enabling customers to summarize text and multimodal content, including images and documents, directly in Snowflake.
With this release, Restricted Session Scope (RSS) is generally available. An RSS is a privilege ceiling that limits what an agent can do on behalf of a user. It intersects with RBAC and never grants privileges the user doesn't already have through their roles.
You can now use data lineage to see which data a Cortex Agent can reach through the tools declared in its specification. A semantic view that a Cortex Analyst tool references, and a Cortex Search service that a Cortex Search tool references, both appear upstream of the agent.
Providers can now add the code_execution and code_toolset_all sandbox tools to Cortex Agents created by a Snowflake Native App.
The Cortex Agents asynchronous API is now generally available on AWS and Microsoft Azure.
Snowflake recommends transitioning to Cortex Agents , which supports every Cortex Analyst capability with higher answer quality. Snowflake adds new natural language analytics capabilities to Cortex Agents.
The Cortex Agents Coding Agent is now generally available. Add the code_toolset_all tool type to a Cortex Agent, and Snowflake provisions and manages a sandbox backed by the same runtime that powers Snowflake CoCo: bash, file read/write/edit, grep, glob, web search, SQL execution, and skills.
Cortex Agent and Cortex Analyst evaluations now let you control the versioned configuration used for scoring: Version targeting pins the configuration under evaluation, not the agent behavior that gets evaluated: each run invokes the agent again and scores a newly generated trace.
The Cortex Agent code execution tool is now in public preview. When you enable the tool on an agent, the agent can generate and run Python in a secure, isolated sandbox to process data, perform calculations, and produce visualizations.
Power BI ingestion for Semantic View Autopilot is now generally available. You can upload a .pbit or .pbix file to Semantic View Autopilot and have a semantic view created automatically from your existing data model, preserving the DAX measures, table relationships, and column descriptions you’ve already built in Power BI and making them available to Cortex...
Snowflake Cortex AI_MULTI_EMBED is now generally available for semantic video search with support for large video files up to 6 GB using the TwelveLabs Marengo Embed 3.0 model, enabling organizations to index and search production-grade video content directly in Snowflake.
Cortex Agents and MCP servers in Snowflake Native Apps are now generally available. Providers can create the following objects in an app’s setup script: Apps can also use inter-app communication so that one app’s agent calls another app’s agent or MCP server.
With this preview release, the AI Security tab in the Trust Center includes an AI Agents Inventory section. This section provides a consolidated inventory of the Cortex Agents in your account, filtered by one or more categories, so that you can identify which agents to assess or investigate for better governance and security posture.
Snowflake Cortex AI_TRANSCRIBE now supports M4A audio files and MOV (QuickTime) video files, including M4A audio encoded with Apple Lossless (ALAC). M4A and MOV are common across Apple devices, screen and video recording tools, and media production workflows, so you can now transcribe more of your existing recordings directly in Snowflake without converting...
With this preview release, AI_EXTRACT and AI_PARSE_DOCUMENT support documents stored on stages that use client-side or server-side encryption, including in accounts that use PrivateLink or other network policies that restrict public network access to stages. For more information, see the following topics:
With this release, we are pleased to announce the general availability of the SHOW CORTEX BASE MODELS command. Use this command to list the Cortex Base Models available in your account, along with each model’s lifecycle status, regional availability, legacy date, and end-of-life date.
Snowflake Cortex AI_TRANSCRIBE now supports AAC (Advanced Audio Coding) input, both as a standalone audio file and as the audio track inside a supported video file.
Snowflake Cortex AI_TRANSLATE now supports translation of up to 100,000 input tokens, enabling organizations to translate large documents and long-form text directly in Snowflake.
Snowflake now supports resource budgets to control credit spending for Cortex Search services. Resource budgets let you define a monthly spending limit and take automated actions, such as revoking access or suspending the service, when spending exceeds your limits. For more information, see Resource budgets for Cortex Search .
You can now use an SQL query (rather than a physical table) as a logical table in a semantic view. This lets you define logical tables based on joins, transformations, or any SQL expression without creating a separate view first.
When defining a semantic view , you can now specify sample values and enum indicators in the CREATE SEMANTIC VIEW command. These features were previously available only in the YAML specification and are now supported in DDL as well. For more information, see Adding sample values and enum indicators .
AI functions, Cortex Agents, Cortex Search and Cortex Analyst inside Snowflake.
Other Snowflake products: Cortex Code, Snowflake, Snowflake Horizon Catalog, Snowflake Native Apps and Marketplace, Snowflake Openflow
| Release-notes page | Site | Feed |
|---|---|---|
| All release notes docs.snowflake.com | docs.snowflake.com | None |
| Behavior changes docs.snowflake.com | docs.snowflake.com | None |
| Clients, drivers and libraries docs.snowflake.com | docs.snowflake.com | None |
| Feature updates 2026 docs.snowflake.com | docs.snowflake.com | None |
| JDBC driver docs.snowflake.com | docs.snowflake.com | None |
| Kafka connector docs.snowflake.com | docs.snowflake.com | None |
| Native SDK for Connectors docs.snowflake.com | docs.snowflake.com | None |
| ODBC driver docs.snowflake.com | docs.snowflake.com | None |
| Preview features docs.snowflake.com | docs.snowflake.com | None |
| Python connector docs.snowflake.com | docs.snowflake.com | None |
| Release notes overview docs.snowflake.com | docs.snowflake.com | None |
| Server release notes docs.snowflake.com | docs.snowflake.com | None |
| Snowpark Connect for Spark docs.snowflake.com | docs.snowflake.com | None |
| Snowpark Library for Python releases github.com | github.com | RSS |
| Snowpark Python docs.snowflake.com | docs.snowflake.com | None |
| Snowpipe Streaming SDK docs.snowflake.com | docs.snowflake.com | None |
| Spark connector docs.snowflake.com | docs.snowflake.com | None |
Releasesby Glyph