Announcing Gemini 3 Within Snowflake Cortex AI: A Deeper Google Cloud Collaboration to Help Customers Build Smarter, Faster and More Secure AI
We are thrilled to announce the expanded availability of Google’s latest Gemini models, Gemini 3 Pro and Gemini 2.5 Flash, running natively within Snowflake Cortex AI. For customers, this means faster access to industry-leading AI, without moving or copying sensitive enterprise data. With Gemini 3, customers also gain access to state-of-the-art multimodal reasoning with unprecedented depth and nuance, enabling agents that can understand subtle context, intent and domain-specific complexity with far less prompting.
This launch deepens our strategic relationship with Google Cloud and strengthens our shared commitment to helping enterprises build generative and agentic AI applications directly where their governed data already lives. This results in a simpler architecture, lower cost and higher trust in agentic AI applications. By combining Snowflake’s governed data foundation with Google’s most intelligent Gemini generation yet, customers can now build agents that are not only powerful but reliably grounded in their enterprise data.
As part of this extended collaboration, Snowflake and Google Cloud are aligning more closely across product innovation and go-to-market efforts. Together, we’re enabling customers globally to build, deploy and scale secure enterprise AI, while giving them the freedom to work across clouds with comprehensive data governance capabilities.
Gemini arrives to Cortex AI
Starting today, Gemini 3 Pro and Gemini 2.5 Flash are available in public preview through Snowflake Cortex AI Functions, with support coming soon to Snowflake Intelligence and Snowflake Cortex Agents. The Gemini models join other frontier models from Anthropic, Meta, OpenAI and more, all running inside Snowflake’s secure and governed perimeter. Gemini 3 Pro brings Google’s latest advances in multimodal understanding, long-context reasoning and agentic capabilities directly into Snowflake, so customers can unlock deeper insights and more capable agents without impacting security or governance controls.
What customers can achieve with Gemini in Snowflake
The Gemini models are extensible to a variety of use cases, depending on your need:
Gemini 3 excels at multimodal reasoning, long-context analysis, advanced tool use and complex agentic workflows. It’s a powerful fit for Snowflake customers building production-grade AI agents that operate securely across structured, unstructured and multimodal enterprise data. This makes it ideal for customers building enterprise-grade AI agents that must reason deeply over financial records, logs, product data, documents and more, securely at scale. Built on Google’s most intelligent Gemini generation to date, Gemini 3 Pro delivers state-of-the-art reasoning with unprecedented depth and nuance, significantly outperforming previous generations across major reasoning, math, multimodal and factual accuracy benchmarks, giving enterprises higher confidence in their most complex, business-critical use cases.
Gemini 2.5 Flash is optimized for speed and cost efficiency. It delivers high-quality results across tasks such as summarization, chat, data extraction and captioning, making it ideal for high-volume workloads that demand both performance and value. It’s perfect for high-volume scenarios such as call summarization, invoice processing and rapid customer service interactions, without slowing down the service.
And because Gemini runs natively in Snowflake, customers can use Gemini models across all supported clouds via cross-region inference. To get started, enable GCP_US or ANY_REGION in your Snowflake admin settings to run Gemini models across all clouds via cross-region inference.
Using Gemini with Cortex AI Functions to unlock enterprise insights faster
Cortex AI Functions make it easy to build cost-efficient, scalable AI pipelines across multimodal enterprise data using familiar SQL. This eliminates complex orchestration and reduces operational overhead, enabling analysts and developers to apply natural language instructions directly to governed data sets. When combined with Gemini 3’s advanced reasoning, customers can move beyond basic Q&A to nuanced analysis, planning and decision support that better captures the real-world complexity of their businesses. For example, using AI_COMPLETE function, users can leverage state-of-the-art models like Gemini 3 Pro:
SELECT
SNOWFLAKE.CORTEX.AI_COMPLETE(
'gemini-3-pro',
PROMPT(
'Review the following financial filing and summarize key revenue trends, margin changes, and risk factors: {0}',
filing_text
)
) AS financial_summary
FROM financial_documents;For customers, this means lower cost, faster development cycles and AI that runs on secure, trusted data already in Snowflake.
Snowflake on Google Cloud: Expanding access, choice and data sovereignty for customers
Our collaboration with Google Cloud continues to accelerate worldwide. Over the past year and beyond, Snowflake has expanded or will expand to new Google Cloud regions to help customers power their most critical workloads:
Kingdom of Saudi Arabia: Aligned with the Kingdom's Vision 2030 strategy to provide secure, faster, cost-efficient data solutions to support Saudi Arabia’s digital transformation.
Melbourne (coming early this year): Increasing flexibility for enterprises across Australia and New Zealand to choose the cloud provider that best suits their operational and compliance needs.
These regional launches broaden access to Snowflake’s AI Data Cloud, enabling customers to operate globally with local control.
A deeper relationship that simplifies customer adoption of AI
Our expanded relationship aligns Google Cloud and Snowflake across products and go-to-market motions to deliver a simpler, more unified customer experience. This includes:
Integrations across Google Cloud’s suite of services, from generative AI with Gemini to machine learning with Google Cloud’s Vertex AI platform and interoperability with BigQuery
Aligned account teams for smoother solution design and support
Google Cloud Marketplace availability so customers can apply committed cloud spend to Snowflake with streamlined procurement
Faster procurement and deployment for AI and data workloads
This gives customers less friction, fewer contracts to manage and faster results.
“Google has long set the standard for innovation and large-scale infrastructure, remaining at the forefront of the generative AI revolution. By combining Google Cloud’s industry leadership with Snowflake’s ability to bring AI directly to enterprise data, we’re empowering customers to move faster, innovate more freely and redefine what’s possible through data and AI.”
Christian Kleinerman
Looking ahead: Delivering more value through secure, production-ready AI
Gemini 3 Pro and Gemini 2.5 Flash are now available in preview within Cortex AI Functions, with integrations coming soon to Snowflake Intelligence and Cortex Agents. As Gemini 3 continues to advance the state of the art in multimodal, agentic and long-horizon reasoning, Snowflake will keep bringing these capabilities directly to governed enterprise data so customers can safely operationalize the next generation of enterprise intelligent agents.
Snowflake and Google Cloud are enabling enterprises to operationalize AI safely, cost effectively and at scale, turning their data into intelligent, automated systems that drive real business impact.
Learn more about our relationship and get started with Gemini in Snowflake.
Forward-looking statements
This article contains forward-looking statements, including about our future product offerings, and are not commitments to deliver any product offerings. Actual results and offerings may differ and are subject to known and unknown risk and uncertainties. See our latest 10-Q for more information.
Authors
