Snowflake Revenue Crossed $1.2 Billion in Q1 FY2027

Snowflake reported in its Q1 FY2027 earnings (February through April 2026, results published June 5, 2026) that product revenue reached $1.21 billion, a 25 percent year-over-year increase from $966 million in Q1 FY2026 and the first quarter in Snowflake’s history in which product revenue exceeded $1.2 billion — a milestone that reflects the expanding commercial adoption of Snowflake Cortex AI, the built-in AI inference and machine learning capability layer embedded directly within the Snowflake Data Cloud platform, which allows enterprise data teams to run large language model inference, vector search, document extraction, and text classification against data already resident in Snowflake’s cloud data warehouse without requiring data movement to an external AI service API or the management of separate AI infrastructure outside the Snowflake environment. Snowflake’s Q1 FY2027 investor filings show total revenue of $1.24 billion (including professional services), remaining performance obligations (RPO) of $5.8 billion — up 28 percent year over year from $4.5 billion at the end of Q1 FY2026 — providing contracted forward revenue visibility that reflects the multi-year commit structure of Snowflake’s capacity-based pricing model, where enterprise customers pre-purchase Snowflake credits at volume discount rates and consume those credits as they run queries, pipelines, and AI workloads against their Snowflake environment. Net revenue retention remained above 120 percent in Q1 FY2027, reflecting the consumption expansion dynamic of Snowflake’s pricing model: as enterprises add new data pipelines, expand AI workload usage through Cortex, and onboard additional business units to shared Snowflake environments, the per-customer credit consumption grows without requiring a new contract negotiation — the expansion occurs organically as additional usage events are billed against the pre-purchased credit pool, generating NRR above 100 percent from customers who are growing their data and AI workload volume faster than their credit commitments anticipated. Snowflake’s customer count reached 11,200 enterprises at the end of Q1 FY2027, with customers generating more than $1 million in trailing 12-month product revenue numbering 590 — up from 485 in Q1 FY2026 — with the large-customer cohort generating approximately 60 percent of total product revenue and representing the enterprise data platform buyers who have standardised their cloud data warehouse, data engineering, and increasingly their AI inference workloads on Snowflake’s unified environment rather than managing separate systems for storage, compute, and AI. Microsoft Intelligent Cloud’s Q3 FY2026 revenue crossing $30 billion frames Snowflake’s hyperscaler positioning: Snowflake’s largest cloud infrastructure partnership is with AWS (where approximately 50 percent of Snowflake’s customer workloads run), with Azure (25 percent) and Google Cloud (25 percent) composing the remainder — a multi-cloud neutrality that positions Snowflake as the enterprise data platform that runs on whichever cloud the enterprise’s primary workloads are hosted on rather than requiring a specific cloud commitment, distinguishing Snowflake’s value proposition from Google BigQuery (which runs only on Google Cloud), Amazon Redshift (AWS-only), and Microsoft Fabric (Azure-only) by delivering data portability and cross-cloud data sharing that enterprises with multi-cloud architectures require. Palantir’s revenue crossing $1 billion in Q1 2026 establishes the enterprise AI data platform structural comparison: where Palantir’s AIP builds the AI agent reasoning layer on top of the Palantir Ontology — a semantic graph that abstracts enterprise data into addressable objects — Snowflake’s Cortex AI builds the AI inference layer directly within the SQL query execution engine that Snowflake’s enterprise customers already use for analytics, allowing data analysts to call LLM inference functions (COMPLETE, EMBED_TEXT, CLASSIFY_TEXT, EXTRACT_ANSWER) from within SQL queries against tables already in Snowflake, without requiring the data team to learn a new programming paradigm or manage separate AI model infrastructure outside their existing Snowflake data stack. IBM watsonx’s software revenue crossing $7 billion in Q2 2026 provides the open lakehouse architecture comparison: IBM’s watsonx.data deploys an open-format data lakehouse (Apache Iceberg, Presto, Spark) that positions enterprise data in vendor-neutral open table formats accessible by any query engine, while Snowflake’s proprietary storage format and compute infrastructure (the micro-partitioned columnar storage that Snowflake’s query engine optimises) provide higher query performance within the Snowflake environment but at the cost of the format-level vendor neutrality that IBM’s open-standard approach preserves — with the two vendors targeting different enterprise architectural philosophies (Snowflake for enterprises that prioritise performance and managed infrastructure, IBM watsonx.data for enterprises that prioritise open-standard interoperability and regulatory auditability over query optimisation). Snowflake’s non-GAAP product gross margin reached 76 percent in Q1 FY2027, consistent with prior quarters, reflecting the mature cloud infrastructure efficiency of the Snowflake service — where the per-credit infrastructure cost (the AWS, Azure, and Google Cloud compute and storage that Snowflake purchases wholesale and resells to customers in the form of Snowflake credits) is largely fixed at the negotiated hyperscaler rates and does not increase with the higher AI workload intensity that Cortex AI customers add, because AI inference in Snowflake Cortex runs on the same general-purpose compute infrastructure as SQL query execution rather than requiring dedicated GPU resources with materially different unit economics from the CPU-based data warehouse compute that Snowflake’s standard pricing covers. Snowflake’s adjusted operating income reached $248 million in Q1 FY2027, a 20 percent non-GAAP operating margin, with free cash flow of $345 million — reflecting the operating leverage that the consumption-based business model generates as customer credit consumption grows above the original contract committed level (which Snowflake books as incremental overage revenue at full contribution margin) and as Snowflake’s sales and marketing cost per dollar of new ARR improves as the AI-related word-of-mouth and partner-driven pipeline within the existing 11,200-customer base reduces the direct sales effort required for customer expansion relative to new customer acquisition.

Snowflake Cortex AI — the in-platform AI capability suite comprising LLM inference (calling GPT-4o, Claude Sonnet, Llama 3, Mistral Large from within SQL via the COMPLETE function), embedding generation (converting text to vector representations for semantic search via EMBED_TEXT), and Cortex Analyst (a natural language to SQL interface that allows business users to query Snowflake data in plain English without writing SQL) — reached 4,500 enterprise customers with active Cortex AI usage in Q1 FY2027, up from 1,200 in Q1 FY2026, with Cortex Analyst representing the highest adoption velocity among new Cortex features because it eliminates the SQL writing barrier that prevents non-technical business stakeholders from accessing the Snowflake data that their organisation’s data engineering team has prepared and loaded. Snowflake Arctic — the enterprise-optimised large language model that Snowflake Research released in April 2024 as an open-source model (available under an Apache 2.0 licence on Hugging Face) trained on a hybrid data mixture of 350 billion instruction tokens emphasising coding and SQL generation tasks — continued its commercial deployment in FY2027 as the default low-cost inference option within Snowflake Cortex, with Snowflake offering Arctic inference at significantly lower per-token pricing than the frontier model options (GPT-4o, Claude Sonnet) within the COMPLETE function, allowing enterprises running high-volume text classification, data extraction, and SQL generation workloads to optimise their Cortex AI cost structure by routing simpler AI tasks to Arctic and complex reasoning tasks to frontier models based on task requirements. Snowflake’s Document AI — the Cortex AI module that extracts structured data fields from unstructured documents (PDFs, Word documents, images containing text) stored in Snowflake’s file storage layer using multimodal AI vision models — added 850 enterprise customers in Q1 FY2027 and contributed meaningfully to Cortex AI’s expansion beyond pure SQL analytics use cases into the document intelligence workflows (contract data extraction, invoice processing, regulatory filing analysis) that represent adjacent AI automation opportunities within the Snowflake customer base without requiring those customers to implement a separate document processing service outside their Snowflake environment. Datadog’s AI observability reaching 3,000 enterprise customers establishes the observability layer for Snowflake AI workloads: Datadog’s Snowflake integration — monitoring query latency, credit consumption, and warehouse utilisation through Datadog’s infrastructure monitoring platform — is among the most widely deployed Datadog integrations, and Datadog’s LLM Observability product adds Cortex AI inference monitoring (token consumption per COMPLETE call, model latency distribution, prompt evaluation scores) to the existing Snowflake infrastructure monitoring that enterprise data platform teams already run through Datadog. Gartner’s 2026 Magic Quadrant for Cloud Database Management Systems positions Snowflake as a Leader for the fifth consecutive year, with Gartner’s evaluation citing Snowflake’s unified data and AI platform architecture (combining data warehouse, data lake, data engineering, and AI inference in a single governance-controlled environment) and the Snowflake Marketplace (4,000-plus data and application listings where enterprise customers can access third-party data products and share data securely with partners through Snowflake’s native data sharing without data copying) as the strongest competitive differentiators against Google BigQuery (whose serverless architecture offers lower operational overhead but less multi-cloud flexibility), Amazon Redshift (deeply integrated with AWS services but locked to the AWS cloud), and the emerging Databricks Data Intelligence Platform (private company, approximately $3 billion ARR, growing at 50 percent annually, positioning its lakehouse architecture as the AI-first alternative to Snowflake’s warehouse-native AI). Bloomberg Technology’s coverage of Snowflake’s Q1 FY2027 $1.2 billion product revenue milestone contextualised the result against the Databricks competitive narrative: Bloomberg noted that while Databricks’ 50 percent ARR growth rate outpaces Snowflake’s 25 percent and Databricks’ Unity Catalog data governance and MLflow experiment tracking represent competitive advantages in the data engineering and machine learning workload segment, Snowflake’s superior SQL analytics performance (consistently ranking highest on TPC-DS benchmark tests at enterprise data volumes), the Snowflake Marketplace’s data sharing network effects, and the enterprise procurement preference for Snowflake’s predictable credit-based billing over Databricks’ per-cluster hour variable cost model sustain Snowflake’s revenue scale above Databricks’ at the FY2027 timeframe. Snowflake’s FY2027 product revenue guidance of $5.25 to $5.27 billion — implying approximately 24 percent year-over-year growth — reflects management’s confidence that Cortex AI customer expansion from 4,500 to a projected 9,000-plus customers by end of FY2027, the Document AI adoption ramp, and the Cortex Analyst business-user SQL replacement capability will sustain the mid-20s product revenue growth rate that the $1.2 billion Q1 FY2027 milestone demonstrates as operational at full Snowflake customer base scale.

What Snowflake Cortex AI Reaching 4,500 Enterprise Customers Signals About In-Data-Platform AI Inference Adoption

Snowflake Cortex AI reaching 4,500 enterprise customers with active usage in Q1 FY2027 — growing from 1,200 customers just four quarters earlier, a 275 percent increase achieved without requiring those customers to sign new contracts, establish new vendor relationships, or migrate data to a new platform — signals that the in-data-platform AI inference model (where AI capabilities are embedded within the data platform the enterprise already uses rather than requiring a separate AI service integration) represents the path of least enterprise AI adoption resistance for the large segment of enterprise data teams that have standardised their analytics and data engineering workflows on a specific cloud data warehouse and whose adoption of AI capabilities is gated more by integration complexity and data governance risk than by AI capability interest. The Cortex AI adoption trajectory’s implication for enterprise AI platform strategy is that the distribution advantage of existing enterprise data platform relationships (Snowflake’s 11,200 existing customers, all of whom have data in Snowflake and established data governance policies for that data) generates faster AI capability adoption velocity than standalone AI platform vendors competing for new enterprise relationships, because the Cortex AI adoption journey for an existing Snowflake customer requires only adding a COMPLETE or EMBED_TEXT function call to an existing SQL workflow rather than the new vendor evaluation, security review, data transfer agreement negotiation, and model deployment that adopting an external AI service from scratch requires. Snowflake’s 4,500 Cortex AI customers in Q1 FY2027 — representing 40 percent of the total 11,200 Snowflake customer base having tried at least one Cortex AI capability — and the RPO of $5.8 billion suggesting three-plus years of contracted committed revenue at the current product revenue run rate together establish Snowflake’s commercial position in the enterprise AI data infrastructure market as structurally durable even as standalone AI platforms (Databricks, Palantir, IBM watsonx) and hyperscaler AI services (Azure OpenAI, Vertex AI, Bedrock) compete for the AI workload spend that Cortex AI’s in-platform positioning captures from within the existing Snowflake data estate rather than requiring competitive displacement of an existing relationship.