Andy Jassy
President and Chief Executive Officer at Amazon.com
Thanks, Dave. Today, we're reporting $148 billion in revenue, up 11% year-over-year, excluding the impact from foreign exchange rates. Operating income was $14.7 billion, up 91% year-over-year, and trailing 12-month free cash flow adjusted for equipment finance leases was $51.4 billion, up 664% or $44.7 billion year-over-year. As I think about what matters to customers long-term and, therefore, to Amazon, there's a lot to like about what we're seeing. Starting with AWS, year-over-year revenue growth accelerated again from 17.2% in Q1 to 18.8% in Q2. We're continuing to see three macro trends drive AWS growth. First, companies have completed the significant majority of their cost optimization efforts and are focused again on new efforts.
Second, companies are spending their energy again on modernizing their infrastructure and moving from on-premises infrastructure to the cloud. This modernization enables builders to save money, innovate at a more rapid clip, and drive productivity in most companies' scarcest resources, developers. This is the flip I've talked about in the past, where the vast majority of global IT spend today is on-premises, and we expect that to keep inverting over time. With the broadest functionality, the strongest security and operational performance, and the deepest partner ecosystem, AWS continues to be customers' partner of choice and the biggest beneficiary of this flip from on-premises to the cloud.
And third, builders and companies of all sizes are excited about leveraging AI. Our AI business continues to grow dramatically with a multi-billion dollar revenue run rate despite it being such early days, but we can see in our results and conversations with customers that our unique approach and offerings are resonating with customers. At the heart of this strategy is a firmly held belief, which we've had since the beginning of AWS, that there is not one tool to rule the world. People don't want just one database option or one analytics choice or one container type. Developers and companies not only reject it, but are suspicious of it. They want multiple options for flexibility and to use the best tool for each job to be done.
The same is true in AI. You saw this several years ago when some companies tried to argue that TensorFlow would be the only machine learning framework that mattered, and then PyTorch and others overtook it. The same one model or one-chip approach dominated the earliest moments of the generative AI boom, but we have a lot of data to suggest this is not what customers want here either. And our AWS team is determined to deliver choice and options for customers. You can see this philosophy in the primitive building blocks we're building at all three layers of the Gen AI stack. At the bottom layer, which is for those building generative AI models themselves, the cost to compute for training and inference is critical, especially as models get to scale.
We have a deep partnership with NVIDIA and the broadest selection of NVIDIA instances available, but we've heard loud and clear from customers that they relish better price performance. It's why we've invested in our own custom silicon in Trainium for training and Inferentia for inference. And the second versions of those chips, with Trainium coming later this year, are very compelling in price performance. We're seeing significant demand for these chips. These model builders also desire services that make it much easier to manage the data, construct the models, experiment, deploy to production, and achieve high-quality performance, all while saving considerable time and money. That's what Amazon SageMaker does so well, including its most recently launched feature called HyperPods that changes the game and networking performance for large models. And we're increasingly seeing model builders standardized on SageMaker.
While many teams will build their own models, lots of others will leverage somebody else's frontier model, customize it with their own data and seek a service that provides broad model selection and great generative AI capabilities. This is what we think of as the middle layer, what Amazon Bedrock does and why Bedrock has tens of thousands of companies using it already. Bedrock has the largest selection of models, the best generative AI capabilities in critical areas like model evaluation, guardrails, RAG and agenting, and then makes it easy to switch between different model types and model sizes. Bedrock has recently added Anthropic's Claude 3.5 models, which are the best performing models on the planet. Meta's new Llama 3.1 models and Mistral's new Large two models. And Lama's and Mistral's impressive performance benchmarks in open nature are quite compelling to our customers as well.
At the application or top layer, we're continuing to see strong adoption of Amazon Q, the most capable generative AI powered assistant for software development and to leverage your own data. Q has the highest known score and acceptance rate for code suggestions, but it does a lot more than provide code suggestions. It tests code, outperforms all other publicly benchmarkable competitors on caching security vulnerabilities and leads all software development assistants on connecting multiple steps together and applying automatic action. It also saves development teams time and money on the muck nobody likes to talk about.
For instance, when companies decide to upgrade from one version of a framework to another, it takes development teams many months, sometimes years, burning valuable opportunity costs and churning developers who hate this tedious, though important work. With Q's code transformation capabilities, Amazon has migrated over 30,000 Java JDK applications in a few months, saving the company $260 million and 4,500 developer years compared to what it would have otherwise cost. That's a game changer. And think about how this Q transformation capability might evolve to address other elusive but highly desired migrations. During the past 18 months, AWS has launched more than twice as many machine learning and generative AI features into general availability than all the other major cloud providers combined. This team is cooking, but we're not close to being done adding capabilities for our customer's usage base.
For perspective, we've added over $2 billion in advertising revenue year-over-year and generated more than $50 billion in revenue in the trailing 12 months. Sponsored ads drive the majority of our advertising revenue today and we see further opportunity there. Even with this growth, it's important to realize we're at the very beginning of what's possible in our video advertising. In May, we made our first appearance at the upfronts and were encouraged by the agency and advertiser feedback on the differentiated value we offer across our content, reach, signals, and ad tech. With ads and prime video, the exciting opportunity for brands is the ability to directly connect advertising that's traditionally been focused on driving awareness, as is the case for TV, to a business outcome, like product sales or subscription signups.
We're able to do that through our measurement and ad tech, so brands can continually improve the relevance and performance of their ads. While ads have become the norm in streaming video, we aim to have meaningfully fewer ads than linear TV and other streaming TV providers. And of course, for customers preferring an ad-free experience, we offer that option for an additional $2.99 a month. In our stores business, we saw growth of 9% year-over-year in the North America segment and 10% year-over-year in the international segment. A few notes on our North America revenue growth rate. First, last quarter's leap day added about 100 basis points of year-over-year growth. Second, we're seeing lower average selling prices, or ASPs, right now because customers continue to trade down on price when they can.
More discretionary higher ticket items, like computers or electronics or TVs, are growing faster for us than what we see elsewhere in the industry, but more slowly than we see in a more robust economy. And our continued faster delivery speed is earning us more of our customers' everyday essentials business. Third, our seller fees are a little lower than expected given the behavior changes we've seen from our latest fee changes. While some of these issues compress short-term revenue, we generally like these trends. While consumers are being careful on price, our North America unit growth is meaningfully outpacing our sales growth as our continued work on selection, low prices, and delivery is resonating. So far this year, our speed of delivery for prime customers has been faster than ever before, with more than 5 billion units arriving the same day or next day.
As more customers experience our fast delivery, they look to Amazon for more of their shopping needs, and the continued acceleration of our everyday essentials business is an example of this phenomenon. On seller fees, lowering apparel fees has spurred substantial year-over-year unit growth in apparel. And the incentive we've given sellers to send their items to multiple Amazon inbound facilities so they can save money where they save us effort and money is getting more traction than we even hoped. These collective developments will benefit customers in the form of better selection, lower prices, and faster delivery speed.
There's no shortage of ideas aimed in improving the experience for stores' customers. For instance, we're adding even more value to Prime, recently introducing free restaurant delivery in many of our geos, expanding Amazon's PharmacyRx pass to Medicare members. This is a benefit that gives subscribers all you can consume access to the most common generic medications for just $5 a month, and offering a grocery subscription to help save on grocery items when shopping our U.S. and UK fresh stores.
As we pursue these initiatives, we remain focused on lowering our cost to serve. We have a number of opportunities to further reduce costs, including expanding our use of automation and robotics, further building out our same-day facility network, and regionalizing our inbound network. With more optimal inbound inventory placement, we expect to enable faster speeds, consolidate more orders in one box, and reduce inventory transfers once items reach a fulfillment center.
These cost improvements won't happen in one quarter or one fell swoop. They take technology and process innovation with a lot of outstanding execution, but we see a path to continuing to lower our cost to serve, which as we've discussed in the past, has very meaningful value for customers in our business. As we lower our cost to serve, we can add more low ASP selection that we can support economically, which coupled with our fast delivery, puts Amazon in the consideration set for increasingly more shopping needs for customers.
A few other comments about areas in which we're investing. We remain very bullish on the medium to long-term impact of AI in every business we know and can imagine. The progress may not be one straight line for companies. Generative AI especially is quite iterative, and companies have to build muscle around the best way to solve actual customer problems. But we see so much potential to change customer experiences. We see it in how our generative AI-powered shopping assistant Rufus is helping customers make better shopping decisions. We see it in how our AI features that allow customers to simulate trying apparel items or changing the buying experience.
We see it in how our generative AI listing tool is enabling sellers to create new selection with a line or two of text versus the many forms previously required. We see it in our fulfillment centers across North America where we're rolling out Project Private Investigator, which uses a combination of generative AI and computer vision to uncover defects before products reach customers. We see it in how our generative AI is helping our customers discover new music and video. We see it in how it's making Alexa smarter. And we see it in how our custom silicon and services like SageMaker and Bedrock are helping both our internal teams and many thousands of external companies reinvent their customer experiences and businesses. We are investing a lot across the board in AI, and we'll keep doing so as we like what we're seeing and what we see ahead of us.
We also continue to like the progress in Prime Video. Our storytelling is resonating with our hundreds of millions of monthly viewers worldwide, and the 62 Emmy nominations Amazon MGM Studios recently received is another supporting data point. We recently debuted titles like Fallout, the second most watched original title ever for Prime Video, The Idea of You, which attracted nearly 50 million viewers worldwide in the first two weeks on Prime Video, and Season 4, The Boys, which reached number one on Prime Video in 165 countries in its opening two weeks. And we continue to see momentum in live sports. We recently announced 11-year landmark deals with the NBA and the WNBA. When combined with our original films and shows, partner streaming services, licensed content, and rent or buy titles, Prime Video continues to evolve into the best destination for streaming video.
And for Project Kuiper, our low-Earth orbit satellite constellation, we're accelerating satellite manufacturing at our facility in Kirkland, Washington. We've announced a distribution agreement with Vrio, who distributes direct TV Latin America and Sky Brazil to offer Project Kuiper's satellite broadband network to residential customers across seven countries in South America, and we continue to feel significant demand for the service from enterprise and government entities. We expect to start shipping production satellites late this year and continue to believe this could be a very large business for us. I could go on, but we'll stop here. There's a lot to feel optimistic about over the next several years and the team collectively remains focused on continuing to invent and deliver for our customers in the business.
With that, I'll turn it over to Brian for a financial update.
