Introducing OpenClaw on DigitalOcean: Safe Defaults without Managing Infrastructure
OpenClaw is a popular open-source framework for building personal AI assistants connected to services and messaging platforms such as Telegram, WhatsApp, Discord, and Slack. Whether you are experimenting with a single assistant or scaling to a fleet of always-on agents, DigitalOcean provides a safer production-ready foundation with the operational model that fits your needs.
1-Click Deploy OpenClaw on a Droplet
Available now starting at just $12/month, making it faster and more affordable than ever to run always-on, agentic AI in a security-hardened cloud environment. We’ve also recently improved the experience with a new one-command update script (/opt/update-openclaw.sh) to keep your instance current, alongside a more stable npm-based installation and a refined setup wizard. Deploy now →
OpenClaw on App Platform
Best when you want elastic scaling, simpler operations, and predictable costs as you grow from one agent to many without managing infrastructure. Deploy agents as code, push updates via Git with zero downtime, and run private, stateful assistants with security-hardened defaults. Clear instance-based pricing makes it easy to scale capacity without unexpected cost spikes. Get started →
Now Available: Opus 4.6, Sonnet 4.6, and other models on Gradient™ AI Platform
Gradient AI Platform supports leading text, code, image, and audio models from OpenAI, Anthropic, fal, and more, all running inside your DigitalOcean environment. One control plane, no separate vendor accounts, and predictable billing alongside your existing services so you can build faster with managed inference and safe defaults. Access models via API or the DigitalOcean Cloud Console (media generation models via API only).
Available models include:
- Text & Code: OpenAI GPT-5.2, GPT-5.2 Pro, GPT-5.1-Codex-Max;
- Anthropic Haiku 4.5, Claude Opus 4.6, and Sonnet 4.6
- Image: fal Stable Diffusion XL Fast; fal FLUX.1 (schnell)
- Audio: fal Stable Audio; ElevenLabs TTS Multilingual v2
Explore available models →
Now Available: Prompt caching for these Anthropic models
Use prompt caching with serverless inference chat completion to cache context and use it in future requests. When you send a request, if part of the input is already cached, you are charged a lower price for those cached tokens, and at the standard price for the remaining input tokens. This significantly reduces the cost for inference.
We support prompt caching for the following Anthropic models:
- Claude Sonnet 4
- Claude Opus 4
- Claude Opus 4.1
- Claude Sonnet 4.5
- Claude Opus 4.5
Now Available: GPU Droplets Powered by AMD InstinctTM MI350X GPUs
High-performance GPU Droplets powered by AMD Instinct™ MI350X are now available, offering the high throughput and low latency needed to tackle complex AI workloads. Learn more about the release.
New Features on App Platform: Supabase and native .NET buildpack support
Deploy a production-ready Supabase backend on App Platform with one click, complete with managed PostgreSQL, authentication, REST API, file storage, and real-time capabilities. With our native .Net buildpack support, you can deploy your .NET applications directly from a Git repository without writing or maintaining Dockerfiles.
