Earnings/Recap
DOCNDigitalOcean Holdings, Inc.

Earnings Recap — Q2 FY2026

CY Q3 2026 · Reported August 4, 2026 · Beat 7 of last 7 quarters

The Buildout is a website built on agentic AI tools, currently in beta, and could have factual errors.

Go to the full DigitalOcean Holdings, Inc. company page →
What this means for the buildout

DigitalOcean's accelerating growth and rising revenue per megawatt underscore the shift from bare-metal GPU rental to full-stack AI inference platforms. The company's ability to secure capacity ahead of demand and land large commitments signals continued strength in the AI infrastructure buildout, particularly for inference-heavy workloads. Its focus on open-weight models and software attach could pressure pure-play Neoclouds and inference providers to differentiate beyond raw capacity.

Results vs consensus
EstimateActualvs est
Revenue$279M$281M+0.8%beat
EPS$0.26$0.45+72.9%beat
What was said

DigitalOcean delivered Q2 revenue of $281M, up 29% YoY, with record incremental ARR of $93M. AI customer ARR reached $234M, growing 212% YoY, with 85% coming from non-bare-metal services. The inference engine, launched in late April, attracted over 6,000 customers with token volume up 30x in 60 days, and open-weight models grew from ~15% to ~75% of token volume. The company also strengthened its balance sheet by retiring $472M of convertible notes and secured its first 9-figure annual revenue commitment, lifting RPO to $894M.

Key metrics
Revenue
$281M
Up ~29% YoY, above guidance; more than double the growth rate from Q2 last year
Incremental ARR
$93M
Record quarterly incremental ARR, nearly triple the same quarter last year
AI Customer ARR
$234M
Up 212% YoY; 85% from inference services and core cloud, not bare metal
Adjusted EBITDA Margin
40%
Adjusted EBITDA of $114M; strong profitability alongside accelerating growth
Remaining Performance Obligations
$894M
Up more than 12x YoY with 3.7-year average life; includes first 9-figure annual commitment
Management outlook

Management raised full-year 2026 revenue growth guidance to approximately 30% (from ~26% previously) and now expects to exit Q4 2026 at 35% or more growth. They reiterated confidence in 50%+ revenue growth for 2027, citing incremental committed capacity and strong demand. Q3 2026 revenue guidance is $304M–$307M (32%–34% YoY) with adjusted EBITDA margins of 38%–39%. They also raised adjusted free cash flow margin guidance to 11%–13% for the full year. Management emphasized disciplined capacity deployment, with 20 MW of additional capacity secured and total committed capacity now ~155 MW, and highlighted the emerging AI-native flywheel driving higher-margin, stickier services.

From the call

We are not a GPU rental business. We are a full stack cloud platform that AI native companies depend on to build, run and scale their production AI software.

on Differentiation from bare-metal Neoclouds

We are not a GPU rental business. We are a full stack cloud platform that AI native companies depend on to build, run and scale their production AI software.

on Differentiation from bare-metal Neoclouds

We expect the incremental ARR that we get per megawatt to increase over time.

on Revenue per megawatt trajectory

What analysts asked

How are you meeting the demands of larger scaled customers and managing operational puts and takes to get megawatts online at the right time?

Paddy highlighted the company's track record of running a global cloud business and serving over 500,000 customers. He noted the pace of innovation (over 80 releases since April) and the addition of forward-deployed engineering to support large customers. Matt added that they work with top data center operators and leading chip manufacturers, and have managed to turn up capacity on time despite industry challenges.

What was the impact of pricing on Q2 revenue growth and the updated outlook? Also, where do you expect net leverage to be at year-end and any specific free cash flow guidance including leases?

Matt said the ~30% list price increase on GPU fleets had a modest impact on Q2 but is baked into the 2026 guide, contributing to the higher exit rate. He noted pro forma net leverage of 0.7x after the equitization, well below the 4x guideline. He confirmed the company will be free cash flow positive on any metric in 2026, with adjusted FCF margin of 11%–13%.

Can you elaborate on the 9-figure deals signed this quarter, the adoption of the 5-layer stack, and the monetization roadmap? Also, what are Kevin Van Gundy's early plans for go-to-market?

Paddy said over 70% of AI customers added this year at significant scale are already using core cloud products. He emphasized that agentic workloads require more than just GPUs or tokens, driving attach across databases, storage, and orchestration. On go-to-market, he said the focus is on landing high-quality AI-native workloads with forward-deployed engineering, not scaling sales investment yet.

Potential supply chain impact
CRWVDigitalOcean's emphasis on non-bare-metal AI services and higher ARR per megawatt could pressure CoreWeave's pure GPU rental model, especially as open-weight models reduce the need for massive training clusters.
AMZNDigitalOcean's inference engine and open-weight model support could attract AI-native customers seeking lower cost and more flexibility than hyperscaler offerings, potentially impacting AWS's AI workload share.
MSFTDigitalOcean's focus on open-weight models and cost-efficient inference could compete with Azure's AI services, particularly among developers and AI-native startups.
GOOGLDigitalOcean's AI-native cloud and inference engine may attract workloads that would otherwise go to Google Cloud, especially those prioritizing open models and unit economics.
ORCLDigitalOcean's integrated platform and software attach could challenge Oracle's AI infrastructure offerings, particularly in the mid-market and AI-native segment.
AKAMDigitalOcean's AI-native cloud and inference services could intensify competition with Akamai's cloud computing (Linode) for developer and AI-native workloads.