Integration

DigitalOcean Serverless Inference integration

Run DigitalOcean-hosted and third-party foundation models (OpenAI GPT, Anthropic Claude, Llama, DeepSeek, Qwen and more) without managing infrastructure: chat completions, the Responses and Anthropic-style Messages APIs, text embeddings, image generation, text-to-speech, text-to-video jobs and TypeSafe System One typed evaluations; list the models your key can use. Every generation is billed to your DigitalOcean account (Inference and Agents balance or prepayment) at the model's rate. Not exposed: streaming, fal async models, Agent Inference and Batch Inference. Authentication is a model access key (or a DigitalOcean personal access token) sent as a Bearer token.

Connect DigitalOcean Serverless Inference How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in DigitalOcean Serverless Inference.Built from the 11 DigitalOcean Serverless Inference tools below, on the connection you add once.

What you can do with DigitalOcean Serverless Inference

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

Video, 2 operations

create_videoCreate videoPOSTText-to-video job, billed for real ($0.60 per video on wan2-2-t2v-a14b). Get the MP4 with download_video within 2 hours. Waits for the job and returns the finished result; gives up after 10 minutes.
get_videoGet videoGETReturns the state of a video generation job (status processing, then completed or failed, with error details on failure). Then download the MP4 with download_video.

Completion, 1 operation

create_chat_completionCreate chat completionPOSTSends a chat conversation (messages with roles) to a model and returns its reply in choices[0].message, OpenAI-compatible. Billed for real per input and output token at the model's DigitalOcean rate (prepaid balance only).

Content, 1 operation

download_videoDownload videoGETDownloads the MP4 of a completed video generation job as a file. Results expire 2 hours after the job completes.

Embedding, 1 operation

create_embeddingsCreate embeddingsPOSTTurns a text into an embedding vector (data[0].embedding) for semantic search, RAG, clustering or similarity. OpenAI-compatible; unknown body fields are rejected. Billed per input token at the model's DigitalOcean rate.

Generation, 1 operation

generate_imageGenerate imagePOSTGenerates images from a text prompt and returns them base64-encoded in data[].b64_json. Billed for real per image or token at the model's rate (Stable Diffusion 3.5 Large lists $0.08 per image).

Message, 1 operation

create_messageCreate messagePOSTSends an Anthropic Messages style conversation (user and assistant turns, optional system prompt) and returns the next assistant message in content[]. Billed for real per token at the model's rate (prepaid balance only).

Model, 1 operation

list_modelsList modelsGETLists the models available to your key for serverless inference, with their id (the value to pass as model) and owner.

Response, 1 operation

create_responseCreate responsePOSTGenerates a model response with the Responses API (input text plus optional instructions and tools); the reply is in output[].content. Billed for real per input and output token at the model's DigitalOcean rate (prepaid balance only).

Speech, 1 operation

create_speechCreate speechPOSTConverts text to speech; the audio (mp3 or wav) comes back base64-encoded in data[0].b64_json. Billed for real at the model's rate (Qwen 3 TTS lists $20.00 per 1M character tokens).

Systemone, 1 operation

create_system_one_evaluationCreate system one evaluationPOSTEvaluates a text (state) against named typed questions with TypeSafe's Jev model and returns typed answers (true/false, a choice or a rubric score) with calibrated probabilities, not free text. Billed per use at the model's DigitalOcean rate.

Use DigitalOcean Serverless Inference from Claude Code, Cursor or Codex

Connect DigitalOcean Serverless Inference in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its DigitalOcean Serverless Inference operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect DigitalOcean Serverless Inference

Create your workspace
Sign up free. Nothing to install.
Connect DigitalOcean Serverless Inference once
Add your DigitalOcean Serverless Inference connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.