Expanding Model Choice in Snowflake: Gemini 4 Argon is Now Available on Snowflake

Snowflake customers can now take advantage of Gemini 4 Argon natively in Snowflake’s AI tools—bringing deep reasoning and visual analysis to your governed enterprise data for higher-impact business use cases.  

When we talk to enterprise leaders, one core truth comes up repeatedly: Your competitive advantage lives in your enterprise data, but no single AI model fits every business challenge.

To turn raw enterprise data into measurable outcomes, organizations need an open, multi-model strategy. Whether you are analyzing revenue trends, evaluating campaign performance, or modernizing software engineering workflows, success relies on having the flexibility to choose the right model for the right job—without creating data security risks or pipeline complexity.

That is why we are excited to announce that Gemini 4 Argon is now in Public Preview for Snowflake customers natively within Snowflake Cortex AI. It is available across Cortex AI Functions at launch with Snowflake CoWork, Snowflake CoCo and Cortex Agents coming soon across supported clouds via cross-region inference!  

For teams looking to leverage Google’s latest frontier model, Gemini 4 Argon brings sustained reasoning and multimodal visual context directly to your enterprise data.

How Enterprises Leverage AI on Their Data in Snowflake

Bringing frontier models like Gemini 4 Argon into Snowflake is about unlocking the true potential of your enterprise data estate. Rather than treating AI as an isolated tool, Snowflake allows teams to apply reasoning models directly where enterprise data already lives, governed, and structured.

Here is how organizations are using Snowflake's open AI ecosystem to transform operational data into business momentum:

• Sustained Reasoning Across Complex Data Workflows: Enterprise tasks rarely rely on a single query. With Gemini 4 Argon, teams can execute multi-step investigations—such as tracing financial discrepancies across complex billing tables or using CoCo to refactor legacy codebases and trace multi-file dependencies.

• Connecting Visual Evidence to Business Analysis: In business reporting, textual analysis is only half the story. In CoWork, teams can evaluate complex financial statements, multi-region sales charts, and campaign performance dashboards simultaneously—connecting analysis directly to visual chart evidence to spot trend outliers faster.

• Batch Document Synthesis at Scale: Through model-selectable AI Functions, operations, legal, and support teams can apply reasoning models across thousands of unstructured records, legal filings, and support tickets in parallel to extract key context and generate detailed summaries.

• Unified Governance Without "Shadow AI": Innovation shouldn't mean bypassing security. Because Gemini 4 Argon runs natively inside Snowflake Cortex AI, every call stays within Snowflake’s security and governance perimeter.

Real-World Value: Enterprise AI in Action

The value of enterprise AI comes into focus in everyday workflows. Support teams can automate email triage and ticket classification. Clinicians can use AI-assisted documentation to summarize patient visits. Retailers can turn customer interactions into more personalized shopping experiences. Across these use cases, the foundation is the same: applying AI to relevant enterprise data within a governed environment.

Firework, a video commerce company, illustrates how that foundation can also power a customer-facing product. Firework wanted to bring the personalized attention of an in-store associate to online shopping. Its challenge was not simply generating answers—it was grounding those answers in high-quality information about shoppers’ needs.

Using Snowpark and Cortex AI, Firework aggregated, cleaned and classified thousands of anonymous customer conversations to understand consumer interests and pain points. That data foundation helped power its AI Video Assistant, which answers return-policy questions, summarizes product reviews and offers personalized recommendations. By keeping the end-to-end workflow within Snowflake, Firework reported a simpler process and greater time and cost efficiency—connecting data preparation directly to a more personalized shopping experience.

With Gemini 4 Argon now available, teams can plug Google Cloud's latest reasoning and visual capabilities into the same governed data pipelines to tackle even more complex, multi-step business challenges.

Get Started Today

Enterprise AI value is unlocked through choice, control, and open interoperability. By adding Gemini 4 Argon to Snowflake Cortex AI we are giving Snowflake customers another powerful option to build, iterate, and drive impact on their trusted data.

To get started, visit our Google Cloud Partner Hub to explore our multi-model capabilities.