Skip to main content

One prepaid balance for practical AI work

Use leading AI models without juggling subscriptions.

Start in browser chat, create a website or app, or connect a tool you already use. Prepay once, choose the model for each job, and see what every request consumed.

Try Web Chat — $2 freeSee the model catalogSee how it works

Already have an account? Sign in to your dashboard

No subscription · Exact total shown before checkout · Usage visible from your dashboard

One line of integration, not a migration.

Point your existing OpenAI-compatible or Claude-compatible client at CloudService and keep the SDK you already use. Model access and terms are confirmed per key in your dashboard.

The integration is one line

Fromhttps://api.openai.com/v1
Tohttps://api.cloudservice.services/v1

Same request shape, same SDK. Point base_url here and authenticate with a CloudService key.

Prepaid, not subscribed: load a balance once, spend it against any model in the catalog, and see exactly what each request drew down from your dashboard.

See the full model catalog below for what each model is good for, or sign in to your dashboardfor your account’s current rates and balance.

Four offerings, one CloudService account.

The AI API gateway is detailed for the rest of this page. Cloud services, web solutions, and custom AI agents are scoped and delivered as their own engagements.

AI API gateway

Prepaid credit routed through documented OpenAI-compatible and Claude-compatible endpoints — GPT, Claude, Gemini, Grok, and Qwen models on one balance.

See the AI gateway

Web solution

Marketing sites, web applications, and integrations — scoped, built, and handed over in full.

Start a web solution

Make your own AI agent

Custom AI agents — tool integration, evaluation, and deployment — built around your task.

Build your own AI agent

Start in three steps

  1. 1

    Start free

    Create a trial key with $2 credit—no card.

  2. 2

    Pick a surface

    Open Web Chat, Studio, or a supported client.

  3. 3

    Stay in control

    See model, token, and balance usage in your dashboard.

Open Web Chat

No install required

Web ChatStudioAPI

Choose a surface. Use your CloudService key.

Three surfaces inside the AI API gateway — the first of the four offerings above.

Web Chat

Ask, write, research, and work with the models available to your key in a focused browser conversation.

Learn about Web Chat

Studio

Describe a website or app, generate project files, and review the result in one workspace.

Learn about Studio

API

Connect documented OpenAI-compatible and Claude-compatible clients with customer-visible usage.

Explore the API

Working evidence, not invented testimonials.

Inspect the live catalog, service state, setup instructions, and support path before you pay.

Live statusCurrent service checks and incident historyModel catalogCopyable API IDs and published customer ratesDocumentationManual and automatic setup pathsHuman supportPublic and signed-in support tickets

A simpler way to run the AI work you already have.

Keep the tools you like, use a clear prepaid balance, and move between supported model families without maintaining a separate billing workflow for every surface.

  1. 01

    No monthly subscription

    Prepay what you need and see the exact checkout total before you confirm.

  2. 02

    One account, several surfaces

    Use the same CloudService account across Web Chat, Studio, and supported API clients.

  3. 03

    A key that matches your plan

    Use API-credit keys or a family-specific token allowance, with renewal rules that stay predictable.

  4. 04

    Usage you can see

    Track requests, tokens, credits, and key status from the dashboard instead of guessing.

Turn a plain-English brief into a project you can review.

Describe the website or app you want, let Studio create the project files, then inspect the structure before you continue. It is a focused starting point—not a black box.

  1. 1Describe — Write the outcome in your own words.
  2. 2Generate — Create an editable project structure.
  3. 3Review — Open the generated files and continue refining.
Open Studio
studio / project preview

Prompt

Build a clean customer API dashboard with usage cards, a model picker, and a responsive navigation.

Project plan ready · files can be reviewed before the next step.

generated project

▾ app

▾ components

UsageCard.tsx

ModelPicker.tsx

page.tsx

globals.css

4 files · ready to inspect

Bring the client you already use.

CloudService publishes setup guides for the clients below. Compatibility depends on the client protocol, selected key scope, base URL, and capabilities returned by the authenticated model catalog.

See all configuration guides

Make your first API call in a minute.

Use your CloudService key as a Bearer token with the OpenAI-compatible API. The request below uses the public fable-5 model ID, which requires a Claude-capable or both-scope key.

Credit-mode base URL
https://api.cloudservice.services/v1
Token-budget base URL
https://api.cloudservice.services/token/v1
Authentication
Authorization: Bearer ak_live_…
Open full quick startAuthentication guideGuided setup
First requestPOST /v1/chat/completions
export CLOUDSERVICE_API_KEY="ak_live_..."

curl https://api.cloudservice.services/v1/chat/completions \
  -H "Authorization: Bearer $CLOUDSERVICE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"fable-5","messages":[{"role":"user","content":"Reply with exactly OK"}],"max_tokens":16}'
For a token-budget key, replace /v1 with /token/v1. The CloudService CLI and supported-client setup select the correct base URL for your key automatically.

Using Claude or Claude Code?

CloudService also supports Claude-compatible requests and documented Claude Code setup for Claude-scoped or both-scoped keys.

Claude Code with CloudService

GPT, Claude, Gemini, Grok and Qwen, behind one prepaid key.

Exact public API IDs and self-service availability. Sign in to your dashboard for your account’s current rates and balance.

Open live model catalog

GPT-5.6 Luna

gpt-5.6-luna
GPT-compatible

GPT-5.6 Terra

gpt-5.6-terra
GPT-compatible

GPT-5.6 Sol

gpt-5.6-sol
GPT-compatible

Claude Opus 5

claude-opus-5
Claude-compatible

Fable 5

fable-5
Claude-compatible

GLM 5.2

glm-5.2
CloudService

Gemini 2.5 Flash Lite

gemini-2.5-flash-lite
CloudService

Gemini 3.6 Flash

gemini-3.6-flash
CloudService

Gemini 3 Pro Preview

gemini-3-pro-preview
CloudService

Grok 4.5

grok-4.5
CloudService

Qwen 3.7 Plus

qwen3.7-plus
CloudService
Understand model IDsSee how billing works

Answers before you start.

Is CloudService a SaaS subscription?

No. CloudService is prepaid AI API access with model pricing, usage tracking, and OpenAI- and Claude-compatible gateway endpoints. You buy credit or a token allowance up front — there is no recurring subscription.

Do I need a new SDK?

No. Use OpenAI-compatible SDKs with the matching CloudService base URL, or use native Claude-compatible clients with a Claude-capable key. Chat Completions, Responses, and Anthropic Messages paths are documented separately.

Can I see my usage?

Yes. Check key status, expiry, credit balance, request counts, and token usage from your dashboard at any time.

What is the difference between credit and token keys?

A credit key draws down a USD balance against published model rates. A token key spends a fixed prepaid allowance for a model family. The base URL differs by billing mode and protocol, and the installer selects the matching route from the authenticated key.

More in the documentation or contact support.

Start in Web Chat, create in Studio, or connect your own app.

Use CloudService directly in the browser, or follow the documentation to connect your API key from SDKs, OpenAI-compatible tools, Claude, or Claude Code.

Open Web Chat Open StudioRead docs