Sep 2, 2026 earnings call — AI products drive Snowflake revenue acceleration
Product revenue was $1.49 billion, accelerating to 37% year over year growth.
Revenue $1.5B (est $1.5B) · EPS $0.62 (est $0.45) · Next-quarter revenue outlook $1.6B–$1.6B
Key quotes
"Meanwhile, CoWork continues to see rapid adoption surpassing 9.1 thousand accounts and adding more than 2,000 net new accounts in this quarter alone."
"AI brings new workloads to the platform, extending our reach to new users, and drives greater consumption across the business."
"In Q2, non GAAP operating margin expanded over 400 basis points year over year to reach 15%."
Analyst Q&A 12
Quality of recent growth
Q (Sanjit Singh · Morgan Stanley): He asked whether the acceleration in revenue is high quality. He asked why Coco and Cowork win in new use cases and whether customers use them responsibly.
A: Management said the growth is broad across many customers, not just AI-native firms. They said Coco makes optimization and cost management much easier and that sales teams use this to build trust and win more projects.
"First, I think we see the acceleration come from a very broad swath of customers." — Sridhar Ramaswamy (CEO)
How much growth from new products
Q (Kirk Materne · Evercore): He asked how much of the acceleration comes directly from new AI products versus the core platform. He wanted a sense of the split.
A: Management said their AI products including Coco and Cowork, plus AI functions and the AI gateway, are a large part of the acceleration. They said these AI products contributed approximately half of the acceleration and that faster migrations and other products also helped.
"They contributed approximately half of the acceleration that we are seeing" — Sridhar Ramaswamy (CEO)
Model choice and neutrality
Q (Karl Keirstead · UBS): He asked if Coco mainly uses Frontier models and whether customers are shifting between open-source and closed models. He also asked if model neutrality is a competitive edge.
A: Management said customers are switching models to optimize cost and performance, and Snowflake supports many open-source models with different economics. They said neutrality and automatic routing to the right model for the right task is a clear competitive advantage.
"And as Sridhar said, automatic routing into what is the right model for the right task, So definitely a very strong, advantage for us." — Christian Kleinerman
Regional growth outside North America
Q (Raimo Lenschow · Barclays): He asked how the business is performing in regions like Europe and Asia and whether there is opportunity to expand there.
A: Management said performance is strong across all regions and that their outlook for every region is included in guidance. They emphasized that all regions are operating very well.
"And the outlook for all regions are factored into our guidance." — Brian G. Robins (CFO)
Why AI lifts Snowflake revenue now
Q (Ryan MacWilliams · Wells Fargo): He asked what changed so AI is now accelerating Snowflake revenue: product improvements, better models, or customers being ready.
A: Management said many things combined: products like Cowork let customers get quick, flexible value from data and Coco makes the sales team AI-native and more confident. They pointed to faster time-to-value and improved cohort metrics as the flywheel driving acceleration.
"What products like co work firmly demonstrated was the ability to get really flexible and quick value from data." — Sridhar Ramaswamy (CEO)
Internal use of Coco and CoWork
Q (Matt Hedberg · RBC): He asked how deep Coco use is inside Snowflake and whether CoWork becomes the place for repeatable agent use cases.
A: Management said they use Coco and Cowork across almost every function and business process internally. They said many internal teams in finance and other areas use these tools, with over 150 internal Snowflake-on-Snowflake uses and near 100% adoption in finance.
"Matt, the pattern that we are seeing is we are leveraging Coco and CoWork throughout pretty much every function and every key business process throughout Snowflake." — Christian Kleinerman
Consumption uplift from AI adopters
Q (BofA): He asked how much more AI adopters consume versus non-adopters and whether the effect is durable or just projects pulled forward.
A: Management said they are not ready to share exact uplift numbers but they measure cohort behavior and see noticeable effects as Coco adoption deepens. They said the breadth and depth of new use cases give confidence this is not only pulled-forward projects.
"At this time, we are not ready to share the exact Uplift numbers, but we do measure cohort behavior." — Sridhar Ramaswamy (CEO)
CoWork adoption trends and onboarding
Q (Brent Thill · Jefferies): He asked whether CoWork adoption shows bigger lands, more users, and larger consumption over time after launch.
A: Management said they track penetration carefully from legal signoff to daily active users and have a pipeline for adoption stages. They said they now use in-product guidance and hands-on labs (3 hours) and share best practices globally to drive adoption at scale.
"But more importantly, we also now have a suite of tools ranging from in product you know, guidance within SnowSite to hands on labs that we do for 3 hours with our customers." — Sridhar Ramaswamy (CEO)
Snowflake model strategy (Arctic)
Q (Brad Zelnick · Deutsche Bank): He asked whether Snowflake plans to build its own large frontier model like Arctic or keep focusing on mix-and-match models.
A: Management said they are not training frontier-scale models but continue to develop Arctic-family models for specific, constrained tasks for higher accuracy and efficiency. They said they will mix Frontier closed models, open-weight models, and their own fine-tuned models to pick the right model for each task.
"We have not changed the direction we have been on, which is we are not training models to go get into a frontier type of model." — Christian Kleinerman
New budgets and Frontier program
Q (Aleksandr Zukin · Wolfe): He asked how big the budget opportunity is now that AI opens access to larger strategic priorities and about traction from the Frontier program.
A: Management said AI lowers the distance from data to business value and enables many practical projects across industries. They described the Frontier Engineering program combining industry expertise, platform harnesses, and outcome commitments with examples like Sanofi, and said they often get paid when outcomes are delivered.
"We have talked publicly about working with folks like Sanofi in our Frontier Engineering program." — Sridhar Ramaswamy (CEO)
Apps becoming databases and overlap
Q (Tyler Radke · Citi): He asked whether SaaS apps partnering with LLMs will become databases and how Snowflake should respond or compete.
A: Management said as software gets easier, data and semantics become more important and owning the user experience matters; Coco and Cowork are fundamental to that. They also said Snowflake offers interoperability, customers prefer consolidating data in a central platform, and Snowflake can host analytic and operational applications using hybrid tables and Postgres.
"I have been very, very consistent for now 2+ years in my conviction, in our conviction that owning the user experience is critical." — Sridhar Ramaswamy (CEO)
What raised full-year guidance
Q (JPMorgan): He asked whether the higher full-year guide increase came from core growth or AI products and whether something changed this quarter to raise visibility.
A: Management said they set guidance based on observed behavior up to the call, and that strength came from both AI products and increased business in the core. They said the guidance reflects the combined strength of AI and the underlying core.
"We base our guidance based on observed behavior up until the call that we have." — Brian G. Robins (CFO)