cURL, Python, and TypeScript | 51cowork API Docs

Quick answer

OpenAI SDKs use https://api.51cowork.com/v1. Anthropic SDKs use https://api.51cowork.com. API Keys come from Dashboard and model IDs come from /v1/models.

HTTP: cURL
Python: openai / anthropic
TypeScript: openai / @anthropic-ai/sdk
Credentials: Environment variables

Shared environment

Terminalbash

export COWORK_GATEWAY_URL="https://api.51cowork.com"
export COWORK_API_KEY="<key-from-dashboard>"
export COWORK_MODEL="<model-from-gateway>"

cURL

OpenAI Chat Completionsbash

curl "$COWORK_GATEWAY_URL/v1/chat/completions" \
  -H "Authorization: Bearer $COWORK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "'"$COWORK_MODEL"'",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

Anthropic Messagesbash

curl "$COWORK_GATEWAY_URL/v1/messages" \
  -H "x-api-key: $COWORK_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "'"$COWORK_MODEL"'",
    "max_tokens": 256,
    "messages": [{"role": "user", "content": "Hello"}]
  }'

Python

OpenAI SDK · pip install openaipython

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["COWORK_API_KEY"],
    base_url=os.environ["COWORK_GATEWAY_URL"].rstrip("/") + "/v1",
)
result = client.chat.completions.create(
    model=os.environ["COWORK_MODEL"],
    messages=[{"role": "user", "content": "Hello"}],
)
print(result.choices[0].message.content)

Anthropic SDK · pip install anthropicpython

import os
from anthropic import Anthropic

client = Anthropic(
    api_key=os.environ["COWORK_API_KEY"],
    base_url=os.environ["COWORK_GATEWAY_URL"],
)
message = client.messages.create(
    model=os.environ["COWORK_MODEL"],
    max_tokens=256,
    messages=[{"role": "user", "content": "Hello"}],
)
print(message.content[0].text)

TypeScript / JavaScript

OpenAI SDK · npm install openaitypescript

import OpenAI from 'openai'

const client = new OpenAI({
  apiKey: process.env.COWORK_API_KEY,
  baseURL: `${process.env.COWORK_GATEWAY_URL?.replace(/\/$/, '')}/v1`,
})
const result = await client.chat.completions.create({
  model: process.env.COWORK_MODEL!,
  messages: [{ role: 'user', content: 'Hello' }],
})
console.log(result.choices[0]?.message.content)

Anthropic SDK · npm install @anthropic-ai/sdktypescript

import Anthropic from '@anthropic-ai/sdk'

const client = new Anthropic({
  apiKey: process.env.COWORK_API_KEY,
  baseURL: process.env.COWORK_GATEWAY_URL,
})
const message = await client.messages.create({
  model: process.env.COWORK_MODEL!,
  max_tokens: 256,
  messages: [{ role: 'user', content: 'Hello' }],
})
console.log(message.content)

Streaming with an SDK

Use the SDK streaming API and consume each event or chunk. Configure connection and read timeouts, and abort the upstream request when the user cancels.

Protocol smoke testbash

curl -N "$COWORK_GATEWAY_URL/v1/chat/completions" \
  -H "Authorization: Bearer $COWORK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "'"$COWORK_MODEL"'",
    "stream": true,
    "messages": [{"role": "user", "content": "Count to five"}]
  }'

Do not buffer the entire response before forwarding it.
Do not log event content, prompts, or full Keys by default.
Retry disconnects cautiously to avoid duplicate calls and cost.

Before production

Inject the Key through a Secret Manager or deployment platform.
Set timeouts, cancellation, and bounded retries.
Create a business request ID and correlate it with Dashboard request_id.
Apply sensible input length and content boundaries.
Monitor balance and error rate, and replenish before exhaustion.