Quickstart

Connect your tools in 2 minutes

CloudService is OpenAI- and Anthropic-compatible. Point any client at the CloudService base URL, drop in your key, and send the same requests you already know. Pick your client below.

Which base URL?

OpenAI-compatible clients end in /v1. Native Anthropic clients use the origin only and append /v1/messages themselves. Token-pack keys add a /token segment.

Which model ID?

Always use an exact ID from your key's authenticated catalog. Fetch it with GET /v1/models. The samples below use safe defaults you can swap.

curl · OpenAI

OpenAI-compatible Chat Completions over plain HTTP.

API-credit keys use https://api.cloudservice.services/v1. OpenAI token-pack keys use https://api.cloudservice.services/token/v1.

chat.sh

curl "https://api.cloudservice.services/v1/chat/completions" \
  -H "Authorization: Bearer YOUR_CLOUDSERVICE_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.4-mini",
    "messages": [{"role":"user","content":"Reply with exactly OK"}],
    "max_tokens": 16
  }'

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

cURL reference →

curl · Anthropic

Native Anthropic Messages over plain HTTP.

Native Anthropic uses the origin only. API-credit keys use https://api.cloudservice.services; Claude token-pack keys use https://api.cloudservice.services/token. The client appends /v1/messages itself.

messages.sh

curl "https://api.cloudservice.services/v1/messages" \
  -H "Authorization: Bearer YOUR_CLOUDSERVICE_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-haiku-4-5-20251001",
    "max_tokens": 16,
    "messages": [{"role":"user","content":"Reply with exactly OK"}]
  }'

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

Anthropic Messages reference →

OpenAI SDK

Point the official OpenAI SDK at the CloudService base URL.

Set the base URL to https://api.cloudservice.services/v1 for API credit, or https://api.cloudservice.services/token/v1 for an OpenAI token pack.

Python

client.py

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_CLOUDSERVICE_KEY",
    base_url="https://api.cloudservice.services/v1",
)

completion = client.chat.completions.create(
    model="gpt-5.4-mini",
    messages=[{"role": "user", "content": "Reply with exactly OK"}],
)

print(completion.choices[0].message.content)

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

Node

client.mjs

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "YOUR_CLOUDSERVICE_KEY",
  baseURL: "https://api.cloudservice.services/v1",
});

const completion = await client.chat.completions.create({
  model: "gpt-5.4-mini",
  messages: [{ role: "user", content: "Reply with exactly OK" }],
});

console.log(completion.choices[0].message.content);

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

OpenAI SDK reference →

Anthropic SDK

Point the official Anthropic SDK at the CloudService origin.

Set the base URL to https://api.cloudservice.services for API credit, or https://api.cloudservice.services/token for a Claude token pack. Do not append /v1 — the SDK builds /v1/messages itself.

Python

client.py

from anthropic import Anthropic

client = Anthropic(
    api_key="YOUR_CLOUDSERVICE_KEY",
    base_url="https://api.cloudservice.services",
)

message = client.messages.create(
    model="claude-haiku-4-5-20251001",
    max_tokens=16,
    messages=[{"role": "user", "content": "Reply with exactly OK"}],
)

print(message.content[0].text)

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

Node

client.mjs

import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({
  apiKey: "YOUR_CLOUDSERVICE_KEY",
  baseURL: "https://api.cloudservice.services",
});

const message = await client.messages.create({
  model: "claude-haiku-4-5-20251001",
  max_tokens: 16,
  messages: [{ role: "user", content: "Reply with exactly OK" }],
});

console.log(message.content[0].text);

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

Claude Code reference →

VS Code Continue

Continue talks to CloudService over OpenAI-compatible Chat Completions.

Continue uses the Chat Completions transport. Keep useResponsesApi: false — this route is not a Responses or tool-calling endpoint.

Store the key outside the config

~/.continue/.env

CLOUDSERVICE_API_KEY=YOUR_CLOUDSERVICE_KEY

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

Add the model

~/.continue/config.yaml

models:
  - name: CloudService GPT-5.4 mini
    provider: openai
    model: gpt-5.4-mini
    apiBase: https://api.cloudservice.services/v1
    apiKey: ${{ secrets.CLOUDSERVICE_API_KEY }}
    useResponsesApi: false
    roles:
      - chat
      - edit

Continue reference →

Cursor

Add a custom OpenAI-compatible model in Cursor's model settings.

Cursor Settings → Models → add a custom OpenAI-compatible model. Use the custom form for Claude models too, not the built-in Anthropic login.

Custom model fields

text

Base URL:  https://api.cloudservice.services/v1
API key:   YOUR_CLOUDSERVICE_KEY
Model:     gpt-5.6-terra   (or another exact catalog ID, e.g. fable-5)

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

Equivalent settings entry

settings.json

{
  "openai.baseURL": "https://api.cloudservice.services/v1",
  "openai.apiKey": "YOUR_CLOUDSERVICE_KEY"
}

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

Cursor reference →

Claude Code

Route Claude Code through the native Anthropic Messages gateway.

Use the origin only in ANTHROPIC_BASE_URL — https://api.cloudservice.services for API credit, https://api.cloudservice.services/token for a Claude token pack. Never append /v1.

Environment (current shell)

env.sh

export ANTHROPIC_BASE_URL="https://api.cloudservice.services"
export ANTHROPIC_AUTH_TOKEN="YOUR_CLOUDSERVICE_KEY"
export ANTHROPIC_MODEL="claude-opus-5"

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

Persistent settings

~/.claude/settings.json

{
  "env": {
    "ANTHROPIC_BASE_URL": "https://api.cloudservice.services",
    "ANTHROPIC_AUTH_TOKEN": "YOUR_CLOUDSERVICE_KEY",
    "ANTHROPIC_MODEL": "claude-opus-5",
    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "claude-haiku-4-5-20251001",
    "CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY": "1",
    "CLAUDE_CODE_DISABLE_THINKING": "1"
  }
}

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

Claude Code reference →

Codex CLI

Codex connects through the OpenAI-compatible Responses provider.

wire_api = "responses" is required. API-credit keys use https://api.cloudservice.services/v1; OpenAI token-pack keys use https://api.cloudservice.services/token/v1.

Provider config

~/.codex/config.toml

model = "gpt-5.6-terra"
model_provider = "cloudservice"

[model_providers.cloudservice]
name = "CloudService"
base_url = "https://api.cloudservice.services/v1"
wire_api = "responses"
requires_openai_auth = true

printf "%s" "YOUR_CLOUDSERVICE_KEY" | codex login --with-api-key

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

Codex reference →

Cherry Studio

Add CloudService as an OpenAI-compatible custom provider.

Choose the OpenAI Compatible provider. API-credit keys use https://api.cloudservice.services/v1; token-pack keys use https://api.cloudservice.services/token/v1.

Provider fields

text

Provider:  OpenAI Compatible
API host:  https://api.cloudservice.services/v1
API key:   YOUR_CLOUDSERVICE_KEY
Model:     fable-5   (or another exact ID from /v1/models)

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

Cherry Studio reference →

n8n

Wire n8n's AI Agent to CloudService through the OpenAI Chat Model sub-node.

Set the credential Base URL to https://api.cloudservice.services/v1 for API credit, or https://api.cloudservice.services/token/v1 for an OpenAI token pack. Do not append /chat/completions — n8n adds the path itself.

Add an AI Agent node and attach "OpenAI Chat Model" as its Chat Model sub-node. Do not use the standalone OpenAI node (v2): a custom Base URL passes its credential test but 404s at runtime (n8n issue #21651); the AI Agent sub-node handles custom base URLs correctly.
On the sub-node, create a new OpenAI credential with the fields below.
If the model dropdown is empty — it usually is for custom providers — switch the Model field to "By ID" and type an exact catalog ID from the list below.
Claude needs no separate node: type a claude-* ID in the same OpenAI Chat Model node and CloudService translates the request.
For anything the chat node can't do, send the request yourself from an HTTP Request node — settings and body below.

Step 2 · OpenAI credential fields

text

API Key:   YOUR_CLOUDSERVICE_KEY
Base URL:  https://api.cloudservice.services/v1

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

Step 3 · Model IDs to type

text

gpt-5.4-mini
gpt-5.6-sol
claude-opus-5
claude-sonnet-5
claude-haiku-4-5-20251001
deepseek-v4-flash   (requires a DeepSeek-entitled key — other keys get a scope error)

Step 5 · Fallback: HTTP Request node

text

Method:          POST
URL:             https://api.cloudservice.services/v1/chat/completions
Authentication:  Generic Credential Type → Header Auth
Header name:     Authorization
Header value:    Bearer YOUR_CLOUDSERVICE_KEY
Body:            JSON (body.json below)

Replace YOUR_CLOUDSERVICE_KEY with your CloudService key.

body.json

{
  "model": "gpt-5.4-mini",
  "messages": [{ "role": "user", "content": "Reply with exactly OK" }]
}

Need account help or a key? Contact support.