Snowflake releases
Snowflake blog and product updates. New features, breaking changes, security advisories and deprecations - each summarised in plain English and updated continuously.
Tracking 52 Snowflake releases · Updated
- Snowflake Blog blogaisecuritygovernancesnowflakeengineergovernment ·
Snowflake Data-Model-Agent Security Framework for AI Agents
Snowflake introduced a new Data-Model-Agent security framework to address the risks posed by AI agents executing tasks on behalf of businesses. This three-layer approach integrates security into the data, model, and agent layers, offering defense-in-depth for governed AI operations. The framework aims to provide enterprise-grade security, including distinct agent identities, tool governance, and prompt injection defenses, for organizations deploying AI agents.
feature announcement - Snowflake Blog blogsnowflakeengineer ·
Snowflake Expands Operations to Chile
Snowflake has officially launched operations in Chile, recognizing the country's significant economic and technological potential in Latin America. This expansion aims to empower Chilean organizations to accelerate their AI initiatives and transform data into tangible business outcomes. The move is driven by increasing demand for Snowflake's AI Data Cloud in the region, enabling local customers and partners to securely access, share, and derive value from data. This physical presence reinforces Snowflake's commitment to becoming a global company and supporting data-driven decision-making across Latin America.
announcement - Snowflake Blog blogdataaisnowflakeengineer ·
Snowflake and Accenture Partner to Create Agentic Enterprise
Snowflake and Accenture are collaborating to enable an "agentic enterprise" by integrating governed data with intelligent agents capable of making and executing decisions. This partnership aims to address the challenge of deploying AI agents in fragmented environments by providing essential decision context and industry-specific intelligence. The solution leverages Snowflake's data platform as a foundation and Accenture's Context Graph to translate governed data into governed actions across various business functions and industries.
announcement feature - Snowflake Blog blogaisnowflakearchitecthealthcare ·
Snowflake: AI Agents in Healthcare - Key Considerations for Leaders
This article outlines critical questions for healthcare and life sciences leaders adopting agentic AI, moving beyond simple tasks to orchestrating complex workflows. It emphasizes the need for trusted data foundations, robust governance, secure data access, and strategies to mitigate AI hallucinations. The piece advises leaders to assess their organization's readiness for privacy, security, and auditability when deploying these powerful AI agents.
announcement - Snowflake Blog blogaigovernancesnowflakeengineer ·
Snowflake Adds Support for Agentic Resource Discovery Specification
Snowflake is announcing support for the Agentic Resource Discovery (ARD) Specification, an open protocol for cataloging and discovering AI agents and tools across an enterprise. This aims to solve the problem of finding available AI capabilities and makes agents discoverable across different interfaces like Snowflake CoWork and Microsoft Copilot. The specification allows clients to describe a task in natural language and receive ranked matching agents with their descriptions and endpoints, while maintaining data access control with the agent provider.
feature announcement - Snowflake Blog blogdatasnowflakepreviewengineer ·
Snowflake Integrates Postgres for Unified Apps, Analytics, and AI
Snowflake is introducing Snowflake Postgres, a new solution designed to bridge the gap between transactional applications and analytical workloads, eliminating the need for complex ETL pipelines. Announced features include data mirroring for real-time replication and 'Postgres for your data lake' for flexible data synchronization with open formats like Iceberg. These capabilities aim to provide businesses with consistent, up-to-date data for AI and real-time applications, impacting teams managing both OLTP and OLAP systems.
feature announcement - Snowflake Blog blogaisnowflakeengineer ·
Snowflake Apps Now Support Full-Stack Web Apps and AI Code Generation
Snowflake Apps now include support for full-stack web applications, allowing developers to build and deploy them directly on the Snowflake platform. This feature aims to accelerate app development by eliminating infrastructure provisioning and configuration, integrating security, and leveraging AI-powered tools for coding and deployment. Available in public preview, Snowflake Apps are designed for engineers and architects working with data-intensive applications on Snowflake.
feature announcement - Snowflake Blog blogaiawssnowflakegaengineer ·
Snowflake Adaptive Compute Achieves General Availability on AWS
Snowflake Adaptive Compute is now generally available on AWS, offering a workload-aware approach to infrastructure that aims to balance performance, scalability, and ease of use for AI and dynamic data workloads. This feature aims to boost throughput, accelerate time to insight, and free up engineering resources by automating infrastructure management. It is now available in expanded global regions and offers superior price-performance compared to previous warehouse generations.
feature announcement - Snowflake Blog blogdatasnowflakeengineer ·
Snowflake AIM: Unified AI Platform for Data Migration and Modernization
Snowflake announced Snowflake AIM, a new AI-powered platform designed to accelerate enterprise data migration and modernization efforts. This unified solution combines existing tools like SnowConvert AI and Snowpark Migration Accelerator to automate assessment, code conversion, and migration for various workloads. AIM supports both full modernization to Snowflake-native architectures and rapid virtualization of legacy systems, such as Teradata, offering a phased approach to reduce risk and speed up time-to-value for organizations undergoing data platform transformations.
feature announcement - Snowflake Blog blogdatasnowflakeengineer ·
Snowflake Dynamic Tables: Faster Refreshes and Enhanced Features
Snowflake has released significant performance improvements for Dynamic Tables, achieving up to 2.8x faster refresh times on Gen2 warehouses. New features like QUALIFY row/rank = 1 and PRIMARY KEY RELY constraints enhance incremental processing and data stability. These updates benefit data engineers and architects managing autonomous data pipelines by reducing latency and compute costs, with some features now generally available and others in public preview.
feature patch - Snowflake Blog blogdataaiengineer ·
AI Shifts Data Engineering: Declarative Pipelines and Agentic Future
Data engineering is shifting towards a declarative approach to handle AI's demands and evolving from manual tasks to strategic execution. This allows for more efficient pipeline development, enabling the integration of AI tools like coding agents, which are adept at modern, software-defined data engineering practices. The future role of data engineers will focus on advanced data modeling and system architecture, ensuring data quality and availability for AI and applications, as detailed in a new guide on building resilient data pipelines.
announcement - Snowflake Blog blogaigovernancesnowflakeengineer ·
Snowflake Ventures invests in Jedify for enterprise AI context
Snowflake Ventures is investing in Jedify, an autonomous context graph for enterprise AI, to address challenges with data context and trustworthy answers in AI deployments. This partnership aims to provide AI agents with grounded business context by automating the creation and governance of semantic models on Snowflake. Jedify's Snowflake Native App is currently in private preview, with wider availability expected on Snowflake Marketplace.
announcement feature - Snowflake Blog blogaisnowflakeanalystfinance ·
Financial Services AI Adoption Accelerates with ROI and Agentic Focus
Financial services firms are rapidly moving from AI pilots to tangible business value, with 68% reporting quantifiable positive ROI from generative AI. Agentic AI is emerging as the next frontier, although governance will be key for safe scaling, with 30% of firms already using it in production. Proprietary data is now a foundation for competitive advantage, with 92% of firms training LLMs on their own data. AI is also showing a positive workforce impact, challenging notions of job elimination.
announcement - Snowflake Blog blogaisnowflakeengineerhealthcare ·
Snowflake Summit: AI Agents and CoWork for Healthcare
Snowflake announced new AI agent capabilities and updates to its CoWork AI companion at Snowflake Summit, focusing on transforming healthcare workflows. These advancements aim to move the industry from passive AI to proactive, governed intelligence, enabling personalized insights and automated tasks for knowledge workers. The new features, with many in public preview, are designed to address complex data challenges in regulated healthcare environments and support scalable enterprise AI execution.
feature announcement - Snowflake Blog blogdataaisnowflakeengineer ·
Snowflake Enhances AI Data Engineering with New Pipeline Capabilities
Snowflake has introduced new AI-powered features for data engineering, aiming to make pipeline creation more robust and efficient. These capabilities integrate AI directly into workflows, offering agentic assistance, autonomous pipelines, and improved declarative options for SQL and Python developers. The updates focus on reducing manual effort, accelerating time-to-production, and enhancing governance for all data engineers working within or across lakehouses and Snowflake.
feature announcement - Snowflake Blog blogdatasnowflakepreviewengineer ·
Snowflake Hybrid Tables See Major Performance and Cost Improvements
Snowflake has significantly boosted the performance of Hybrid Tables, offering up to 8x faster point lookups and 10x faster batch writes with 10x lower cost, based on internal benchmarks. These enhancements aim to consolidate transactional and analytical workloads, eliminating the need for separate OLTP databases and reducing complexity for AI applications and workflow state management. The improvements, which will be enabled by default for all Hybrid Tables workloads, also include the elimination of request credit billing, resulting in average cost reductions of 15%.
feature patch announcement - Snowflake Blog blogaisnowflakeengineer ·
Anthropic's Claude Fable 5 Available on Snowflake Cortex AI
Anthropic's latest model, Claude Fable 5, is now available in private preview on Snowflake Cortex AI. This Mythos-class model enhances capabilities for complex, long-running tasks, benefiting data engineers, analysts, and developers. It integrates with Snowflake CoCo, Cortex Agents, Cortex AI Functions, Cortex Inference, and Snowflake CoWork, all within Snowflake's security and governance perimeter.
feature announcement - Snowflake Blog blogdataaisnowflakeengineer ·
Snowflake CoCo AI agent gains Snowpark Python deployment skill
Snowflake introduced a new AI-powered skill for its CoCo coding agent, named snowpark-python, designed to streamline the deployment of Python pipelines to Snowflake. This feature aims to reduce the overhead of dependency management, configuration, and tool context-switching for data engineers, enabling them to deploy local Python code to production-scale workflows with a single prompt. The skill is available now within Snowsight, CLI, and IDE extensions, offering authoring, deployment, and observability assistance.
feature announcement - Snowflake Blog blogaisnowflakegaengineer ·
LGND AI Wins 2026 Snowflake Startup Challenge
LGND AI, a company creating Large Earth Models (LEMs) trained on planetary imagery, has won the 2026 Snowflake Startup Challenge. Their technology aims to make Earth observation data queryable by AI, enabling applications in insurance, government, and AI agents. LGND AI leverages Snowflake's platform, including Snowpipe, Dynamic Tables, and Cortex Search, to process vast amounts of geospatial data and build sophisticated AI applications for understanding the physical world. Runners-up Airrived and Twine Security were also recognized for their AI agent innovations.
announcement - Snowflake Blog blogaisnowflakeengineer ·
Snowflake and OpenAI launch production-ready business-native AI
Snowflake and OpenAI have launched a production-ready integration, making frontier AI models available directly within Snowflake Cortex AI. This partnership aims to empower employees with AI that understands business context, securely and at scale, across AWS, Google Cloud, and Azure. The integration offers unified governance and leverages Snowflake's semantic layer for improved accuracy and reduced costs.
feature announcement