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