Q2 FY2026 Earnings Recap

Amazon.com, Inc. (AMZN)

Reported Jul 30th 2026

News Summary

  • AWS chips (Trainium/Graviton) crossed a $25B annualized run rate growing triple digits YoY, with Trainium securing multi-year, multi-gigawatt commitments from Anthropic and OpenAI; critically, Jassy disclosed Amazon is actively exploring selling Trainium directly to third-party data centers outside the AWS cloud model, a potential new competitive vector for custom silicon.
  • AWS demand is structurally supply-constrained through at least 2028: backlog surged to $496B (up from $364B last quarter, growing triple digits YoY), the lion's share of 2027 capacity is already reserved, and Jassy described enterprise AI adoption as still in early innings — the massive middle segment of production workloads has barely begun inference at scale, framing it as the largest future AI compute demand driver.
  • Amazon hiked 2026 capex guidance 10% to $220B (from $200B) citing rising memory chip costs, with Q2 capex alone hitting $54.2B vs $32.1B a year ago; AWS operating margin simultaneously expanded 650bps YoY to 39.4%, with management attributing gains to custom silicon efficiency, and Jassy noted AI margin trajectory is tracking ahead of the pace core cloud followed at the same maturity stage.
  • AWS revenue grew 37% YoY to $42.2B (vs 31% consensus), its fastest growth in 18 quarters, with operating income of $16.6B up 63% YoY; total company revenue was $200.6B (+20% YoY, beat by ~$4B), operating income $27.5B (+43%), though trailing-12-month free cash flow swung to -$7.6B from +$18.2B a year ago as capex scaled; Q3 guidance of $197-202B revenue and $22.5-26.5B operating income came in below the $204.1B Street revenue consensus due to Prime Day timing shift.
  • AMZN surged ~9-10% in after-hours trading; analyst sentiment was strongly positive with Synovus Trust calling AWS market share concerns 'put to bed' and framing results as proof the AI infrastructure investment thesis is working, while press coverage characterized the quarter as validating Big Tech's $700B+ collective AI spend and soothing overbuilding fears across the sector.

Financial Highlights

MetricQ3 '24Q4 '24Q1 '25Q2 '25Q3 '25Q4 '25Q1 '26Q2 '26
Revenue$158.9B$187.8B$155.7B$167.7B$180.2B$213.4B$181.5B$200.6B
Rev Growth (YoY)11.0%10.5%8.6%13.3%13.4%13.6%16.6%▲ 19.6%
Gross Margin49.0%47.3%50.6%51.8%50.8%48.5%51.8%▲ 52.3%
Operating Margin11.0%11.3%11.8%11.4%9.7%11.7%13.1%▲ 13.7%
EV/Sales12.7x12.6x13.4x14.4x13.4x11.9x12.9x13.5x

Earnings Call Highlights

Amazon's Q2 2026 earnings call was dominated by an unmistakably bullish tone on AWS and AI infrastructure, with management projecting extraordinary long-term ambitions including a potential $1 trillion annual revenue business for AWS. Andy Jassy was unusually specific and confident in defending the ROIC logic of $220 billion in 2026 CapEx, framing it as a multi-decade compounding asset rather than a near-term drag. Analysts probed capacity constraints, margin sustainability, and the Trainium chip strategy, and management's responses were expansive rather than defensive, signaling genuine demand-side conviction. The dominant theme was that AWS is in a demand-constrained, not supply-constrained, growth phase, with AI and core cloud reinforcing each other in ways that are accelerating rather than cannibalizing margins.

  • AWS chips revenue crossed a $25 billion annualized run rate growing triple digits year-over-year, with Trainium securing multi-year, multi-gigawatt commitments from both Anthropic and OpenAI, and Jassy disclosed Amazon is 'actively having conversations and exploring' selling Trainium chips directly to third-party data centers outside of the AWS cloud model.
  • AWS revenue grew 36.7% year-over-year — the fifth consecutive quarter of acceleration and the fastest growth in 18 quarters — with the backlog surging to $496 billion growing triple digits year-over-year, and Jassy stated demand already reserved for 2028 is 'striking,' signaling the capacity constraint extends well beyond 2027.
  • AWS operating margin expanded 650 basis points year-over-year to 39%, with Brian Olsavsky attributing it to capacity optimization and custom silicon efficiency gains, and Jassy noted AI margin trajectory is tracking 'a little ahead' of the pace core cloud followed at the same stage of maturity.
  • Jassy described the enterprise AI adoption curve as 'barbelled' — AI labs and a few breakout apps on one end, cost-avoidance enterprise use cases on the other — with the massive middle segment of existing production workloads barely using inference at scale yet, framing this as the largest future demand driver for AI compute.
  • Agentic infrastructure is becoming a concrete revenue driver: Amazon Bedrock Agents added payments, policies, and web search capabilities; Kiro coding agent tripled in usage quarter-over-quarter; and Amazon Q is now deployed at BMW, ExxonMobil, AstraZeneca, and over a dozen other major enterprises, with Jassy citing agent tool use on CPUs as a specific tailwind for Graviton adoption.

Sources

Generated by Earnings Bot on 2026-07-31 14:02 UTC.