Inference growth accelerates the AI cloud story
- DigitalOcean is shifting from simple cloud hosting toward a full AI Native Cloud for inference and agents.
- Management raised 2026 revenue growth guidance to about 30%, expecting 35% growth by the fourth quarter.
- Inference services grew almost 800% year over year and now drive over 70% of AI revenue.
- The 2027 bet depends on bringing 60 megawatts of new data center capacity online and filling it.
- The stock price reflects high expectations, leaving little room for error on product or margin execution.
A faster AI story, with a harder bar
DigitalOcean used to be known mainly as the simpler cloud for developers and smaller tech companies. That core business still matters. The new thesis is bigger: management wants the company to become the easy cloud for AI inference, which means running AI models after they are trained, and for agentic apps.
The company raised its outlook sharply after the second quarter of 2026. Management now expects about 30% revenue growth in 2026, with the fourth quarter reaching at least 35%, and 50% or more revenue growth in 2027. That 2027 target depends on 60 megawatts of new data center capacity. If demand shows up, this could reset how investors view the business.
The early signs are strong. Inference services grew almost 800% year over year in the second quarter, now representing over 70% of total AI customer revenue. Customers are also shifting rapidly toward open weight models, validating the open platform approach against single model providers.
The bear case centers on execution and margins. Gross margin compressed in early 2026 as the company spent ahead of new data center revenue. Competition is intense, and the stock price already reflects high expectations. The central question is whether the company can build capacity, sell higher level services, and earn good returns.
Simple cloud, paid by usage
DigitalOcean makes money when customers use its cloud platform. Pricing is mainly consumption based and billed monthly, so revenue rises when customers run more apps, store more data, or use more compute. This model is easy for small teams to start with, but it can also move down if customers cut usage.
The company focuses on Digital Native Enterprise customers, or DNE customers. These are users that spend more than $500 in a month. DNE customers made up 67% of revenue in the second quarter of 2026, up from 59% a year earlier. That mix shift matters because bigger customers are more likely to buy databases, Kubernetes, GPUs, inference, and support around their apps.
DigitalOcean still uses a self service sales motion. Developers can sign up without a large sales process. The company adds targeted sales help for larger AI native accounts, including marquee logos. This keeps the model simpler than the hyperscalers, but it means the company must prove it can land larger customers without losing its identity.
The main break point is capital intensity. AI workloads need expensive data center capacity and GPUs. If the company builds too early, margins suffer. If it builds too late, customers may go elsewhere.
Five layers for AI builders
Foundational cloud infrastructure
This includes global infrastructure, 20 data centers, CPU and GPU Droplets, Kubernetes, networking, and storage. It is the base layer customers use to run apps.
Inference Engine
This layer gives customers serverless and dedicated endpoints for AI models, batch processing, and routing for cost and performance. It supports open-source and closed-source models.
Data and Learning Layer
This includes managed MySQL and PostgreSQL plus vector database support. It helps AI apps store normal data and search data by meaning.
Managed Agents Platform
This is for building and running autonomous agents at scale. It is early, but it could become important if agentic apps become a major cloud workload.
AI Middleware and Inference Router
This layer routes workloads across models, regions, and hardware based on real-time trade-offs. It can make the platform more useful than a plain GPU provider.
Core developer services
Managed databases, managed hosting, marketplaces, and developer tools keep the older promise alive: make cloud work easier for smaller teams.
Bigger customers now matter most
DigitalOcean does not report formal operating segments. The mix below uses its Q2 2026 customer categories: DNE customers were 67% of revenue, and developers were the remaining 33%.
What could go wrong
60 megawatts arrive late
High impact · Medium oddsThe 2027 growth target depends on bringing 60 megawatts of new data center capacity online and selling into it. Delays in power, equipment, construction, or customer ramps could push revenue out. That would hurt the strongest part of the current bull case.
AI margins disappoint
High impact · Medium oddsGross margin fell in early 2026 because data center costs came before the related revenue. AI infrastructure can be lower margin than software cloud services. If higher level AI services do not scale, the company may grow fast but earn less per dollar of revenue.
Platform edge fades
High impact · Medium oddsDigitalOcean is trying to be simpler than AWS, Azure, and Google Cloud while offering more than bare metal GPU providers. Competitors are also adding inference platforms and agent tools. The company must keep its product easy, open, and useful enough to avoid price competition.
Startup customers pull back
Medium impact · Medium oddsMany customers are startups and growing tech businesses. These customers can cut cloud usage when venture funding gets tighter or the economy slows. Consumption billing helps customers start small, but it also lets them shrink quickly.
High expectations meet a high price
Medium impact · High oddsThe story has improved, but the market is already giving credit for faster growth. If 2027 guidance slips, or if AI revenue slows, the stock could react sharply. Good companies can still be poor investments if the starting price is too high.
In one breath
What does DigitalOcean do?
DigitalOcean sells cloud infrastructure and developer tools. Customers use it to run apps, databases, websites, Kubernetes, storage, GPUs, and now AI inference and agent workloads.
Why is AI important for DOCN?
AI is now the main growth story. AI Customer ARR reached $170 million in early 2026, and management tied its 2027 target of 50% or more revenue growth to new capacity and AI demand.
How is DigitalOcean different from AWS or Azure?
DigitalOcean tries to be simpler and more predictable for developers and AI-native teams. It also offers open-source options at every layer of its AI stack, which can help customers avoid being locked into one cloud or model provider.
What is the biggest risk for DigitalOcean stock?
The biggest risk is execution. The company must bring new capacity online, fill it with customers, and prove AI workloads can produce strong margins.
