The team that maintains SGLang, the engine serving trillions of tokens a day, brought DeepSeek AI V4 to production on DigitalOcean and AMD. 🤝 RadixArk moved to AMD Instinct™ MI350X GPU Droplets and co-engineered a ~10x throughput gain with the AMD team, from debugging on borrowed hardware to production-grade inference.
DigitalOcean
Software Development
Broomfield, Colorado 169,820 followers
AI-Native Cloud. ☁️
About us
DigitalOcean is the AI-Native Cloud purpose-built for the inference and agentic era. Its five-layer integrated platform—spanning GPU and CPU infrastructure, core cloud, inference, data, and managed agent orchestration—is open throughout with no vendor lock-in, giving builders everything they need to start fast, scale production AI workloads, and improve unit economics. More than 650,000 customers and millions of developers globally trust DigitalOcean to build, ship, and scale their applications.
- Website
-
https://www.digitalocean.com
External link for DigitalOcean
- Industry
- Software Development
- Company size
- 1,001-5,000 employees
- Headquarters
- Broomfield, Colorado
- Type
- Public Company
- Founded
- 2012
- Specialties
- Cloud Computing, Cloud Servers, Virtual Hosting, Cloud Hosting, Cloud Infrastructure, Simple Hosting, and Virtual Servers
Locations
-
Primary
Get directions
105 Edgeview Dr
Broomfield, Colorado 80021, US
Employees at DigitalOcean
Updates
-
DigitalOcean reposted this
AI for Social Good hackathon (hosted by Major League Hacking and DigitalOcean) is in full swing 🧠🌿 So fun seeing teams cook up inspiring ideas to make life better for people who need it! I got so excited that I built a site that uses AI to make financial education more accessible (all built on DigitalOcean ofc). Check it out here: https://lnkd.in/gBAa48Mx 🌊 Thanks for joining Shivani Kumar and James Skelton! And for partnering Alex Gornet and Mary Patricia Siebert! #DigitalOcean #AIforGood
-
Kickoff AMD Advancing AI early. 🍻 We’re hosting a happy hour with AMD, RadixArk, and friends in SF. Valuemaxxing your inference and infra? We'll have lightning talks and live demos from teams doing exactly that. If you're building or backing AI-native companies, you won't want to miss this. 🎟️ https://do.co/4vdCMZv
-
-
OpenAI GPT-5.6 Sol, Terra, and Luna are here. Deploy now through DigitalOcean Serverless Inference. 🚀 Three durable capability tiers, one fully managed platform: Sol for frontier-level reasoning, Terra for balanced everyday production, and Luna for high-throughput apps that need speed at a low cost. All three run with predictable, usage-based pricing and no separate OpenAI account or contract required, so your team moves from model choice to production traffic without adding operational overhead. Deploy today. 🔗 https://do.co/4peQ9Xa
-
Managed Weaviate is now in public preview on DigitalOcean. 🚀 Production-ready vector search without the ops burden: automated backups, upgrades & high availability, starting at $20/month with no per-query or per-dimension surcharges. It runs the unmodified open-source Weaviate engine, so your existing clients work without changes and it sits natively alongside our Data & Learning layer for RAG, agentic workflows, and semantic search. Deploy today. 🔗 https://do.co/4wSK0D1
-
Inference and agentic demand is accelerating on DigitalOcean's AI-Native Cloud. ☁️ We expect the following record results for Q2 2026: 📈 29% expected YoY revenue growth, up from 14% in Q2 2025. 🚀 RPO expected to exceed $800M, up more than 10X year over year. 🤝 Multiple nine-figure annual commitments. We're proud to support the next wave of AI-native & digital-first enterprises scaling on DigitalOcean. 🌊
-
-
The hack for a great summer? Spending it with DigitalOcean and Major League Hacking. 🤝
Still haven't carved out time to build something this summer? 🌞 Spots are filling up for the AI hackathon MLH is hosting with DigitalOcean in San Francisco, July 10-11. It's a low-stakes, high-energy way to test out an idea, learn a new tool, or just remember why building is fun — no experience required. You'll get a full day (plus an evening) to hack, mentorship from experienced builders, free food, and a shot at cash prizes. Past hackers have gone on to launch startups off ideas that started as a weekend project. Apply here: https://luma.com/MLHandDO Tag a founder or builder in SF who should be in the room 👇
-
What works at ten requests fails at a million. 🧠⚙️📈 At #DOdeploy, engineering leads from Workato, Hippocratic AI, and Information Security Media Group (ISMG) got candid about what actually breaks when inference workloads move from prototype to production: agent permission sprawl, P99 latency spikes, and architecture that wasn't built for the traffic headed its way. Workato cut inference costs 67% & sped up time-to-first-token by 77% running on DigitalOcean's AI-Native Cloud, while Hippocratic AI held sub-second latency across 20M+ patient interactions.
-
Anthropic Claude Sonnet 5 is now available on DigitalOcean Serverless Inference. 🚀 It plans, calls tools, and handles long-horizon tasks at near-Opus 4.8 performance on SWE-bench, BrowseComp, and OSWorld-Verified. Deploy it through the OpenAI-compatible API with fully managed infrastructure and usage-based pricing that stays predictable as usage scales, no separate contract, no ops overhead.