DigitalOcean Managed Agents is now in public preview. Run Claude Code, Codex, or your own LangGraph agent in a runtime environment that pauses when idle. Put its tools behind one governed endpoint, and pick from 75+ open and proprietary models. One cloud, one bill. Prompts to get started available in the blog: do.co/4ysh3it

Now available: @nvidia Nemotron™ 3 Diarization on DigitalOcean Serverless Inference. Open-weight streaming diarization that tracks who spoke when, up to 8 speakers in live or recorded audio. Drops in alongside the ASR you already run. Start building: do.co/4AvIpW3

Jev doesn't write text. It returns a typed decision in 70–500ms: is this spam, which queue, what score. @typesafeai's new System One model scored 50M product reviews for about $20. A token-billed LLM would cost thousands. Now available through DigitalOcean Serverless Inference. do.co/4xJjqMw

Join us for the night shift after @AIconference. Your agents aren't failing because of the model. They're failing on runtime, auth, and isolation. Lightning talks + live demos with @box and Bright Data. RSVP today. do.co/4AqzvcA

It's time to move your AI agents to the cloud. Qencode built one for customer support triage. A request comes in, the agent wakes up, reads it, and files the ticket properly. @r_o_connor shows you how to build one just like it. Bring the harness, model, and tools you already use. We run them. DigitalOcean Managed Agents is now in public preview. do.co/4xN8zRM

Our latest Open Intelligence Summit speakers: @NousResearch, @inferact, @OpenRouter (@PeterJ_Walker), @Amplitude_HQ, and @boltdotnew (@EricSimons). Hear about coding agents, inference, and the products built on top. Oct 13 at The Midway in SF. Request a seat: do.co/4ApUvzV

Fork a session and you get three agents on the same repo, in the same environment, trying three approaches. Keep the one that works, kill the other two. Managed Agents is in public preview: do.co/4AumqPt

Your agent calls a tool and gets back 20 JSON records, but it only needed one of them. Where do the other 19 go? Into the context window, of course, and you’ll re-send them on every remaining turn. Use code sandboxes instead, and you use 129,021 tokens instead of 268,438 for the same task. Available now as part of DigitalOcean’s Action Gateway. do.co/4AumqPt

New from @OpenAI, now on DigitalOcean Serverless Inference: GPT-6 Luna: classification and routing at volume GPT-6 Sol: multi-step reasoning and coding Same bill as the agents that call them. Try them now in Inference Engine. do.co/4AvIpW3

Every agent launch this month is about software taking an action. The models underneath were built to talk. Jev from @typesafeai was built to decide. Define the answers up front, get typed values with confidence scores back. Now on Serverless Inference. do.co/4rtI8it

1/ “Sandboxes and agent harnesses should NOT be in the same runtime.” That was one reaction yesterday when we lifted the covers on @digitalocean’s Harness Runtime. A quick 🧵, because the definitions matter, and so do the boundaries for isolation, durability and scale.

The modern AI-native stack has a Rube Goldberg problem. One vendor for the harness. Another for the sandbox. More for models, inference, web search, MCP tools, storage, databases and application hosting. Every seam is infrastructure your team must integrate, secure, observe and scale instead of solving customer problems. In this video, I build and deploy a complete multi-agent Prospect Scout on @digitalocean : → OpenCode running in a secure sandbox → DeepSeek V4 Pro and Jev for research and qualification → Scheduled, recurring agent runs → Web search through Action Gateway → Managed PostgreSQL for persistent data → App Platform for the customer-facing experience All managed from one AI Native Cloud platform. With DigitalOcean Managed Agents, builders get secure harness execution, 75+ models, 16,000 tools, inference, storage, databases and cloud infrastructure in one integrated stack. Build and scale agents not the Rube Goldberg machine around it.

Exa is now officially available on DigitalOcean! Give your AI agents access to Exa web search through the DigitalOcean Action Gateway.

Claude Opus 5.5 from @AnthropicAI is now live on DigitalOcean Serverless Inference. do.co/4y36dis It improves most on agentic coding and long-running tasks, explains what it's doing in plainer language, and costs 20% less per token than Opus 5, all inside your existing DigitalOcean environment.

Action Gateway is the half of Managed Agents that handles tools. 16,000+ of them, from 400+ providers, behind one endpoint. Ask to "check if the server is still alive." The tool is called ping_network_endpoint. With no words in common, keyword search makes that tool hard to find. @r_o_connor explains how Action Gateway finds the right one anyway, 99.3% of the time. Try it in public preview: do.co/3Vp76U3

DigitalOcean now offers Managed Agents and access to 75+ open-source and proprietary models—all in one cloud. DigitalOcean is generally less expensive than many major cloud providers and easier to use. Now combine that simplicity with open models. That’s interesting.

Today we shipped DigitalOcean Managed Agents into public preview. Every era of the cloud has an atomic unit. For twenty years it was the provisioned VM. You came for compute and paid by the hour, working or waiting. Agents don't fit that box. They think in tokens, act in bursts of compute, and need memory (state) that outlives the session. Compute, inference, and data working as one. The compute-first cloud meters the VM. The agent-first cloud meters the work. I've talked to a lot of founders building agents, and I hear the same stack every time: a harness on a laptop, sandboxes from one startup, inference from two more, storage at a hyperscaler, search from a sixth vendor. Six vendors means fifteen seams to secure and operate. Nobody can say what a single run cost. So we built Managed Agents deeply integrated with our Inference Engine. Harness Runtime: every session in its own Firecracker microVM. Pause, resume, checkpoint, fork. Bring Claude Code, Codex CLI, OpenCode, Hermes, or your own image. Action Gateway: one endpoint to 16,000+ governed tools. The agent never sees the key. Pricing that matches the workload: active CPU billed by the second. An hour of agent work that bills $0.126 fully allocated costs about $0.031 with us for a typical agent run (25% active). While the agent waits, you pay zero. Agents think, act, remember, and improve. Now the whole loop runs on one platform. Public preview is live today. Bring your harness and choose your model. We'll handle the rest. #managedagents #agents #AInativecloud #digitalocean #Bringyouragents