A model API for developers

One API Key for Leading AI Models

Call six leading model families through one gateway, connect Claude Code, Codex, OpenCode, Cursor, Cherry Studio, and WorkBuddy, and manage your account, API Keys, and every request in one Dashboard.

Get an API Key Read setup docs

GPT / Claude / Grok / DeepSeek / GLM / Kimi
OpenAI / Anthropic compatible
Request-level usage

Dashboard provides account authentication and API Key management.

/loginAccount accessUse a method enabled in Dashboard

/v1/modelsModel catalogModels available to the current Key

API KeysKey controlCreate and manage in Dashboard

Request levelUsage detailModel, tokens, cost, and status

Supported model families

Leading models through one unified API

51cowork supports the GPT, Claude, Grok, DeepSeek, GLM, and Kimi model families through OpenAI- or Anthropic-compatible request formats.

GPT
Claude
Grok
DeepSeek
GLM
Kimi

Use /v1/models and the current Dashboard configuration for exact model IDs, versions, rates, and availability.

Built for developer workflows

Keep your tools. Change the Base URL and Key.

There is no new workflow to learn. Choose the API format your client already supports, then add the 51cowork gateway, your Key, and a current GPT, Claude, Grok, DeepSeek, GLM, or Kimi model ID.

Connect the tools you use

Follow dedicated setup for Claude Code, Codex, OpenCode, Cursor, Cherry Studio, and WorkBuddy.

Model catalog and current rates

Model IDs come from /v1/models. Rates, balance, and available purchase entry points come from the current Dashboard.

Trace every request

Review the model, tokens, cost, and status recorded for every request in Dashboard.

Two familiar API formats

Use the format your client already speaks

The current public gateway is https://api.51cowork.com. Choose the client-native format, then fetch current model IDs from /v1/models.

Setup guides available

Claude Code
Codex
OpenCode
Cursor
Cherry Studio
WorkBuddy

OpenAI-compatiblePOST /v1/chat/completions

For OpenCode, Cursor, Cherry Studio, and WorkBuddy. Codex uses the Responses endpoint.

Anthropic-compatiblePOST /v1/messages

For Claude Code and workflows built around the Messages API.

Available models and capabilities follow /v1/models and the current Dashboard configuration.

Start in three steps

From account registration to your first API request

Account access and Key management happen in Dashboard. Once you have three configuration values, your existing tool can call 51cowork API.

Register or sign in

Open Dashboard and use an account method currently enabled there.

Create an API Key

Create a Key for model requests and store its full value securely.

Select a model and call

Copy a model ID from /v1/models, configure your tool, and send the first request.

Open Dashboard

Quickstart

Copy three values. Send your first request.

Use https://api.51cowork.com as the gateway, create an API Key in Dashboard, and fetch a model ID from /v1/models. Pick any example below to verify the setup.

Before you start

Create and copy your API Key in Dashboard
Choose a current model from /v1/models
Store the full Key in a secure environment variable

client.config

curl "https://api.51cowork.com/v1/chat/completions" \
  -H "Authorization: Bearer <key-from-dashboard>" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "<model-from-/v1/models>",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

export OPENAI_API_KEY="<key-from-dashboard>"
export OPENAI_BASE_URL="https://api.51cowork.com/v1"

# Start your OpenAI-compatible client with these variables.

export ANTHROPIC_API_KEY="<key-from-dashboard>"
export ANTHROPIC_BASE_URL="https://api.51cowork.com"

# Start your Anthropic-compatible client with these variables.

One dashboard

Keep Keys, runtime settings, and usage in one place

No cross-checking between separate systems. Your account, Keys, runtime configuration, and requests are managed in one dashboard.

Account and purchases

Review the account state, balance, and purchase options currently provided by Dashboard.

API Keys

Create, inspect, and maintain the API Keys available to the current account.

Request history

Filter by date and model, then inspect status, input and output tokens, total tokens, and cost.

Client configuration

Check the current gateway, model list, and setup guides whenever you update a local client.

Runtime billing information

Use the current Dashboard for models, rates, and balance

The static site explains billing boundaries. Dashboard provides the live catalog, rates, balance, and any enabled purchase entry point.

RUNTIME PRICING

Use the current configuration

Dynamic ratesMeasured by model and request

The public site does not freeze model prices, purchase methods, or account limits that may change.

Choose a model from /v1/models
Check current rates in Dashboard
Measure actual model calls
Inspect tokens and cost per request
Use runtime purchase options

Open Dashboard

Confirm before calling

Available models: Returned by /v1/models
Model rates: Current Dashboard configuration
Purchases and balance: Follow the account page

Check the model and rate before calling. Add balance only through an entry point currently shown in Dashboard.

View pricing and billing

Clear runtime status

Verify configuration, requests, and usage in Dashboard

Dashboard runtime configuration determines account capabilities. The public site documents stable setup without copying dynamic account state.

Dashboard provides authentication
Runtime defines models and rates
Usage records each request

Open Dashboard

Setup flowKEY / MODEL

1Register
2Create Key
3Call model
4Review usage

Documentation built for setup

A practical guide for every way you connect

From GPT, Claude, Grok, DeepSeek, GLM, and Kimi model discovery to setup and troubleshooting for Claude Code, Codex, OpenCode, Cursor, Cherry Studio, and WorkBuddy, the docs provide complete steps and copyable examples. The current gateway is https://api.51cowork.com.

Quickstart

Follow the complete path from account creation to your first model request.

Account and authentication

Use the registration and sign-in methods currently enabled in Dashboard.

Rates and purchases

Check current model rates, balance, and purchase options in Dashboard.

API Key management

Create and securely store API Keys for model calls.

Client setup

Configure Claude Code, Codex, OpenCode, Cursor, Cherry Studio, and WorkBuddy.

Usage and troubleshooting

Inspect tokens and cost, then diagnose auth, rate limits, streaming, and retries.

Open full documentation

FAQ

What developers ask before starting

How do I register or sign in?

Open Dashboard and use an authentication method currently enabled there. The public site does not authenticate accounts.

Where do I find rates and purchase options?

Use the model rates, balance, and purchase options currently shown in Dashboard.

Which models are supported?

51cowork supports the GPT, Claude, Grok, DeepSeek, GLM, and Kimi model families. Call /v1/models with the current API Key and use an exact model ID from the response.

Do I sign in again on the public site?

No. The public site stores no session; registration, sign-in, and API Key management stay in Dashboard.

How should I store an API Key?

Keep the full Key in a server secret or local environment variable. Never commit it or expose it in browser code.

Where do I inspect requests?

Use the Dashboard Usage page to review model, tokens, cost, and status.

Which Base URL should I use?

The current public gateway is https://api.51cowork.com. OpenAI SDK uses the /v1 Base URL; Anthropic SDK and direct HTTP calls use the gateway root. See the setup docs for exact client configuration.

Start building

Connect your AI tools to 51cowork API

Get an API Key