DigitalOcean Holdings, Inc.

What they do

DigitalOcean operates an agentic inference cloud that combines GPU infrastructure, a full-stack cloud, model-first inference workflows and an agentic experience layer. It sells IaaS (Droplet virtual machines, storage, networking), PaaS/SaaS (Managed Hosting, Managed Database, Managed Kubernetes, Marketplace) and its Gradient AI Agentic Cloud, which includes GPU Droplets, Bare Metal GPUs, LLMs and Gradient AI Agents. Revenue is generated mainly from customer usage of the platform, largely consumption-based and month-to-month, with a growing number of committed contracts.

Revenue drivers

• Digital Native Enterprise (DNE) customers — Customers spending more than $500 per month were approximately 22,000 as of June 30, 2026 and produced 67% of revenue in the three months ended June 30, 2026, up from about 59% a year earlier.
• AI Customer ARR — AI customer annual run-rate revenue was $234 million as of June 30, 2026, up 212% year over year from $75 million, with management stating 85% of AI customer ARR now comes from inference and core cloud rather than bare metal.
• Large committed customers — Revenue from $1M+ customers grew 214% year over year and represented 23% of total revenue in the quarter; $500K+ customers represented 26% and $100K+ customers 35% of total revenue.
• Committed backlog (RPO) — Remaining performance obligation was $894 million at June 30, 2026, up from $71 million a year earlier, of which $366 million is expected to be recognized over the next 12 months.

Recent performance

Second quarter 2026 revenue was $281 million, up 29% year over year, with ARR of $1,125 million, also up 29%. Net income attributable to common stockholders was $35 million, down 4%, and diluted net income per share was $0.29; adjusted EBITDA was $114 million at a 40% margin. Operating cash flow was $110 million and adjusted free cash flow was $61 million. The company added a record $93 million of incremental ARR during the quarter, up 191% year over year.

Strategy

DigitalOcean is positioning itself as an AI-native cloud built for inference and agentic workloads, combining infrastructure, core cloud, inference, data and agents in one open stack. It launched Inference Engine in the quarter and shipped more than 80 product releases since April. It is pursuing large AI-native customers, signing its first nine-figure annual commitments and extending weighted average contract life from 1.6 years to over 3 years. It secured an incremental 20 MW of committed data center capacity expected online in 2027 and 2028, bringing total committed capacity to approximately 155 MW. It also repurchased about $472 million of its 0.00% Convertible Senior Notes due 2030, funded by a concurrent registered direct offering.

Risks

• Customer concentration rising — The top 25 customers made up approximately 20% of revenue in the three months ended June 30, 2026, up from 9% a year earlier, so a few large AI customers now materially affect results.
• Slower margin conversion — Second quarter 2026 net income fell 4% and operating income fell 18% year over year even as revenue grew 29%, while adjusted free cash flow margin declined to 22% from 26%.
• Capacity and capital execution — The business depends on delivering committed GPU and data center capacity on time, with about 155 MW committed and more being pursued, and management cited capacity delivered on or ahead of schedule as a driver of confidence.
• Revenue growth may not persist — The 10-K risk factors state results have fluctuated and are expected to fluctuate, and that prior revenue growth should not be relied on as an indication of future performance.

Outlook

Management raised its 2026 revenue outlook to approximately 30% growth, reaching 35% or more by the fourth quarter of 2026. It also stated conviction in exceeding 50% growth in 2027, citing customer momentum and early product traction, including roughly 30x growth in token consumption among early Inference Engine customers over the last 60 days.