DigitalOcean Gradient™ AI Agentic Cloud

Build, train, and deploy AI agents with the DigitalOcean Gradient™ AI Agentic Cloud:

DigitalOcean Gradient™ AI Platform

Build AI agents on GPU-powered infrastructure using foundation models and resources such as knowledge bases and agent routes.

DigitalOcean Gradient™ AI GPU Droplets

GPU Droplets are VMs with GPUs in a single or 8 GPU configuration. We provide an AI/ML-ready image and 1-Click Models so you can get started without manual setup.

DigitalOcean Gradient™ AI Bare Metal GPUs

DigitalOcean Gradient™ AI Bare Metal GPUs are dedicated, single-tenant servers with eight GPUs per machine. They can run as standalone servers or as part of multi-node clusters.

DigitalOcean Gradient™ AI 1-Click Models

1-Click Models let you deploy third-party generative AI models on DigitalOcean Gradient™ AI GPU Droplets with no additional setup or configuration.

Paperspace

Paperspace is a cloud-based machine learning platform that offers GPU-powered virtual machines and a Kubernetes-based container service.

Bare metal GPUs and GPU Droplets both provide GPU-based compute resources tailored to AI/ML workloads, but they’re each suited for different use cases. Learn more about the difference between bare metal GPUs and GPU Droplets.

Latest Updates

27 February 2026

• We now support prompt caching for the following OpenAI models in serverless inference chat completions and responses API:

  • GPT-5.3-Codex
  • GPT-5.2
  • GPT-5.2 pro
  • GPT-5.1-Codex-Max
  • GPT-5
  • GPT-5 mini
  • GPT-5 nano
  • GPT-4.1
  • GPT-4o
  • GPT-4o mini
  • o1
  • o3
  • o3-mini
  • GPT-image-1

25 February 2026

• The following OpenAI models are now available on DigitalOcean Gradient™ AI Platform for serverless inference and Agent Development Kit:

  • GPT-5.3-Codex

  For more information, see the Available Models page.

• The following Anthropic model is deprecated from DigitalOcean Gradient™ AI Platform:

  • Claude 3 Opus

  Deprecated models have reached end-of-life and are no longer accessible. API or CLI requests to a deprecated model return a model not found error. Migrate to a supported active model to avoid service disruption.

24 February 2026

• You can now use reasoning with serverless inference chat completion. For more information, see Use Reasoning.

For more information, see the full release notes.