

Cortex AI Gateway
#13 v LLM API a gatewaySnowflake · od 28. Juli 2026 (Ankündigung; Public Preview folgt in Kürze) · 9× · naposledy 05. 11. 2026
Cortex AI Gateway is a centralized runtime control plane introduced by Snowflake on July 28, 2026 for governing, monitoring, and securing enterprise AI agents. It manages how first-party agents (e.g., Snowflake CoWork, CoCo) and third-party agents (e.g., Claude Code, Cursor, Amazon Bedrock, Azure AI Foundry) access models, tools, MCP servers, and enterprise systems, keeps an end-to-end log of agent activity, and attributes/limits token-based costs. The product is set to enter public preview soon, with select security and governance features already generally available or in private preview.
Vlastnosti
| Deployment (Self-Hosted/Cloud) | Fully managed cloud service within the Snowflake platform, no self-hosting documented |
| Throughput/Latency | No official throughput/latency figures found; planned 'Intelligent Model Routing' feature (private preview) routes requests partly based on latency |
| Platform | Part of the Snowflake AI Data Cloud / Cortex platform, cloud-based |
| Price | No public price list found for the Gateway itself; general Cortex AI usage is billed via Snowflake AI Credits ($2.00/credit with global routing, $2.20/credit with regional pinning) |
| Protocol Compatibility | Supports 100+ MCP servers (Model Context Protocol) for governance, auth and audit logging |
| Release Date | Announced July 28, 2026 at Black Hat 2026; public preview to follow soon |
| Supported Models/Providers | Model catalog (private preview) including GPT, Gemini, Claude, Grok, Mistral, GLM and more; also governs agents like CoWork, CoCo, Claude Code, Cursor, Amazon Bedrock, Azure AI Foundry |
Další produkty v této kategorii: LLM API a gateway
Zdroje (9)
Company Analysis: Snowflake
Data in the last 90 days supports a stronger operating narrative (Q2 FY27 product revenue acceleration and a raised FY27 product revenue guide to ~$6.07B). (sec.gov) However, compiled valuation metrics remain premium (approx. ~18.6x EV/Sales and ~115x forward P/E), and management disclosures indicate AI workload mix can pressure gross margin near term, increasing the burden of proof for sustained operating leverage. (stockanalysis.com) With the current price not provided (and SNOW trading in USD), a HOLD best fits a data-driven stance: maintain exposure to improving growth/guidance while requiring additional quarters of margin/FCF confirmation before upgrading to BUY.
Summary
Snowflake (NYSE: SNOW) operates a consumption-based cloud data platform (“AI Data Cloud”) that consolidates data warehousing, data lake, data engineering, data sharing/marketplace, and AI/ML workloads on a single governed layer. Its core competency is separating storage and compute while abstracting multi-cloud infrastructure complexity, enabling elastic scaling and cross-functional data access with centralized governance. Monetization is primarily product revenue tied to customer usage (consumption), supplemented by professional services. Market position remains strong in enterprise cloud analytics, with competitive advantages centered on (1) multi-cloud portability, (2) a large and expanding ecosystem (partners, data sharing, marketplace), and (3) governance/security features that are increasingly critical as customers operationalize AI on sensitive datasets. Competitive pressure is most acute from hyperscalers’ native analytics stacks and Databricks; Snowflake’s differentiation is its integrated governance + data sharing + enterprise-grade operational controls. Recent developments (last ~90 days) were dominated by fiscal Q2 2027 results (quarter ended July 31, 2026; reported September 2, 2026). Snowflake reported product revenue of $1.49B (+37% YoY) and raised full-year fiscal 2027 product revenue guidance to $6.07B (implying ~36% YoY growth), citing accelerating momentum and meaningful AI contribution. Management also highlighted rapid product velocity (330+ capabilities GA in 1H FY27) and AI-related customer wins. A notable strategic signal was the expanded five-year collaboration with LSEG (announced September 30, 2026) focused on governed AI use cases in financial services. (sec.gov) Financially, Snowflake remains in an investment phase: GAAP profitability is not the primary near-term metric, while non-GAAP margins and free cash flow are key investor focus areas. In Q2 FY27, non-GAAP product gross margin was reported at ~74% and adjusted free cash flow margin around mid-single digits (per the earnings release coverage), reflecting that AI workloads can be more compute-intensive and may pressure gross margin mix near term even as they support growth. (morningstar.com) Valuation remains demanding versus most software peers. As of mid-November 2026, third-party compiled market data indicated SNOW trading at roughly ~18.6x EV/Sales and ~115x forward P/E (price in USD), implying the equity is priced for sustained high growth and improving operating leverage. (stockanalysis.com) Outlook (short- to medium-term): the raised FY27 product revenue guide supports a constructive demand narrative, but investors should monitor (1) AI-driven gross margin mix, (2) consumption volatility (macro sensitivity), and (3) evidence that AI attach expands net revenue retention without structurally lowering unit economics. (investing.com)
Key Takeaways
- Growth re-accelerated in FY27: Q2 FY27 product revenue was ~$1.49B (+37% YoY) and management raised FY27 product revenue guidance to ~$6.07B. (sec.gov)
- AI is a measurable driver of recent momentum, but it can be more compute-intensive, creating near-term gross margin mix pressure to watch. (investing.com)
- Product velocity and platform breadth are increasing (330+ GA capabilities in 1H FY27), supporting competitive positioning beyond core warehousing. (investors.snowflake.com)
- Strategic enterprise partnerships continue (e.g., expanded five-year collaboration with LSEG), reinforcing Snowflake’s positioning in governed AI for regulated industries. (snowflake.com)
- Valuation remains high (approx. ~18.6x EV/Sales and ~115x forward P/E on compiled market data), so execution against guidance and margin trajectory is critical for returns. (stockanalysis.com)
Action Ideas
Base-case positioning for investors who already own SNOW: the company delivered accelerating product revenue growth and raised FY27 product revenue guidance, which supports the demand narrative. However, valuation multiples remain elevated versus typical software benchmarks, and management commentary/metrics imply AI workload mix can pressure gross margin near term. A HOLD stance is consistent with waiting for clearer evidence that AI-driven growth translates into durable operating leverage and free-cash-flow scaling without sustained unit-economics dilution.
Horizon: 12 mo.
For investors with a 18–24 month horizon and tolerance for valuation risk: Snowflake’s FY27 guide raise to ~$6.07B product revenue and the third consecutive quarter of growth acceleration indicate improving demand traction, with AI cited as a contributor. If Snowflake sustains >30% product revenue growth while stabilizing gross margin and expanding free cash flow, the market can justify premium multiples. This action is most appropriate on pullbacks or for staged entries (risk-managed sizing) given the current multiple profile.
Horizon: 24 mo.
For valuation-sensitive investors or those requiring near-term profitability: SNOW’s compiled forward P/E and EV/Sales remain high, and the company is still balancing growth with profitability. If an investor’s mandate prioritizes near-term earnings/FCF yield, reallocating to lower-multiple software with clearer margin expansion may be warranted. This is a risk-control action rather than a view that fundamentals are deteriorating, given the raised FY27 product revenue guidance.
Horizon: 6 mo.
Contrarian Insights
- • Contrarian (bearish vs. growth narrative): AI-driven acceleration may not be purely additive; if AI workloads are materially more compute-intensive, the near-term trade-off could be structurally lower gross margin mix, making it harder to translate growth into operating leverage at the pace implied by premium valuation. (investors.snowflake.com)
- • Contrarian (bullish vs. ‘hyperscalers will win’ view): Snowflake’s emphasis on governed AI and enterprise-wide collaborations (e.g., LSEG) suggests regulated industries may prefer an independent governance layer across clouds, which can sustain differentiation even as hyperscalers bundle analytics features. (snowflake.com)
Sources (8)
- https://www.sec.gov/Archives/edgar/data/1640147/000164014726000033/fy2027q2earnings.htm
- https://d18rn0p25nwr6d.cloudfront.net/CIK-0001640147/28ed6ad0-102e-4e6e-974f-7122c64fe019.pdf
- https://investors.snowflake.com/financials/quarterly-results/default.aspx?gclsrc=aw.ds®code=snowflakesummit400
- https://www.snowflake.com/en/news/press-releases/lseg-snowflake-expand-collaboration/
- https://www.investing.com/news/stock-market-news/snowflake-lifts-annual-product-revenue-forecast-on-demand-for-cloud-data-platform-4886702
- https://stockanalysis.com/stocks/snow/statistics/
- https://stockanalysis.com/stocks/snow/financials/ratios/
- https://www.ecb.europa.eu/stats/shared/pdf/eurofxref.pdf