Amazon Web Services (AWS) growth sped up for the sixth quarter in a row

Revenue reached $46.0B, adding $3.8B over Q2 alone.

The backlog went from $364B in Q1 to $496B in Q2 and $578B now. Andy Jassy described it as “growing triple digits year over year and well diversified.”

The fair worry with numbers this big is that a few AI labs account for most of it. AWS is saying otherwise.

Looking ahead, Jassy:

“we believe that AWS can grow north of 40% in each of 2027 and in 2028, and be a trillion dollar revenue business in the next 5 to 10 years.”

Bedrock is turning into a multi-model platform

• Fortune 500 spending: ~7x YoY

• Token volume: ~17x YoY

• OpenAI models now on Bedrock, with usage growing quickly

Jassy called Bedrock “the fastest growing multi-billion dollar service in AWS’s history,” and said it could become “as large as EC2.”

His view on models: within two to three years there will be “six to eight models that are about as good as each other,” and “most customers are going to use multiple models.”

Value moves from models to the platform, data and agents around it.

AWS is building the agent layer

AgentCore, AWS’s infrastructure for building and running agents, has tens of thousands of customers.

They deployed 2x more agents in Q3 than in the previous three quarters combined.

Jassy expects “virtually every” enterprise production application to be rewritten for agents, and says “most companies will also use turnkey agentic services,” including AWS’s own.

AI spending is now pulling core cloud along with it

AI revenue is growing at triple digits, and inference is “well over half of AWS’s AI revenue and growing.” At the same time, core compute and storage “grew at their fastest rates in several years.”

The two are connected. Jassy: “customers prefer their production inference to be near the rest of their data and apps,” and AI then drives core demand “because AI uses core for work like reinforcement learning, post-training, and tool usage.”

CFO Brian Olsavsky: “As customers invest in AI, we see a corresponding increase in core consumption.”

Every AI workload drags data, storage and compute behind it. Even mainframes and storage networks, long seen as workloads that would never move, “are moving,” helped by AI-driven migration.

What this means for alliance leaders

• $578B in commitments has to be consumed. That’s a growing pool of spend partners can tap through Marketplace.

• Sell AI together with the migration, data and security work underneath it. That’s where the pull-through lands.

• Expect more agents, including AWS’s own. Differentiation needs to be specific.