OpenAI API Endpoints¶

Complete reference for OpenAI-compatible endpoints in your local Claude Code Proxy.

Base URL¶

http://localhost:8000/openai/v1

All OpenAI-compatible endpoints are prefixed with /openai/v1/.

Authentication¶

No API key required - uses your existing Claude CLI authentication automatically.

Note: OpenAI client libraries require an API key parameter, but any dummy value will work (e.g., "dummy-key").

Chat Completions¶

POST /openai/v1/chat/completions¶

Create a chat completion using OpenAI format, automatically translated to Claude.

Request¶

POST /openai/v1/chat/completions
Content-Type: application/json

{
  "model": "claude-3-5-sonnet-20241022",
  "messages": [
    {
      "role": "user",
      "content": "Hello, how are you?"
    }
  ],
  "max_tokens": 1000,
  "temperature": 0.7,
  "stream": false
}

Request Parameters¶

Parameter	Type	Required	Description
`model`	string	Yes	Claude model to use
`messages`	array	Yes	Array of message objects
`max_tokens`	integer	No	Maximum tokens to generate
`temperature`	float	No	Sampling temperature (0.0-2.0)
`top_p`	float	No	Nucleus sampling parameter
`stream`	boolean	No	Enable streaming responses
`stop`	array/string	No	Stop sequences for completion
`presence_penalty`	float	No	Presence penalty (-2.0 to 2.0)
`frequency_penalty`	float	No	Frequency penalty (-2.0 to 2.0)
`user`	string	No	User identifier

Response¶

{
  "id": "chatcmpl-123",
  "object": "chat.completion",
  "created": 1677652288,
  "model": "claude-3-5-sonnet-20241022",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Hello! I'm doing well, thank you for asking. How can I help you today?"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 12,
    "completion_tokens": 20,
    "total_tokens": 32
  }
}

Streaming Response¶

When stream: true, responses are sent as Server-Sent Events in OpenAI format:

data: {"id":"chatcmpl-123","object":"chat.completion.chunk","created":1677652288,"model":"claude-3-5-sonnet-20241022","choices":[{"index":0,"delta":{"role":"assistant"},"finish_reason":null}]}

data: {"id":"chatcmpl-123","object":"chat.completion.chunk","created":1677652288,"model":"claude-3-5-sonnet-20241022","choices":[{"index":0,"delta":{"content":"Hello"},"finish_reason":null}]}

data: {"id":"chatcmpl-123","object":"chat.completion.chunk","created":1677652288,"model":"claude-3-5-sonnet-20241022","choices":[{"index":0,"delta":{"content":"!"},"finish_reason":null}]}

data: {"id":"chatcmpl-123","object":"chat.completion.chunk","created":1677652288,"model":"claude-3-5-sonnet-20241022","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}

data: [DONE]

Models¶

GET /openai/v1/models¶

List available Claude models in OpenAI format.

Request¶

GET /openai/v1/models

Response¶

{
  "object": "list",
  "data": [
    {
      "id": "claude-3-5-sonnet-20241022",
      "object": "model",
      "created": 1677649963,
      "owned_by": "anthropic"
    },
    {
      "id": "claude-3-5-haiku-20241022",
      "object": "model",
      "created": 1677649963,
      "owned_by": "anthropic"
    }
  ]
}

Parameter Translation¶

The proxy automatically translates OpenAI parameters to Claude equivalents:

OpenAI Parameter	Claude Parameter	Notes
`max_tokens`	`max_tokens`	Direct mapping
`temperature`	`temperature`	Clamped to 0.0-1.0 range
`top_p`	`top_p`	Direct mapping
`stop`	`stop_sequences`	Converted to array format
`presence_penalty`	Not supported	Ignored with warning
`frequency_penalty`	Not supported	Ignored with warning
`user`	Not supported	Ignored

Error Responses¶

Error Format¶

{
  "error": {
    "message": "Invalid request parameters",
    "type": "invalid_request_error",
    "param": "model",
    "code": "invalid_model"
  }
}

Common Error Codes¶

Error Code	Status Code	Description
`invalid_request_error`	400	Request validation failed
`invalid_api_key`	401	Authentication failed
`insufficient_quota`	403	Permission denied
`model_not_found`	404	Model not available
`rate_limit_exceeded`	429	Rate limit exceeded
`internal_server_error`	500	Internal server error
`service_unavailable`	503	Service unavailable

Usage Examples¶

Python with OpenAI SDK¶

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:8000/openai/v1",
    api_key="dummy-key"  # Required but not used
)

response = client.chat.completions.create(
    model="claude-3-5-sonnet-20241022",
    messages=[{"role": "user", "content": "Hello!"}],
    max_tokens=100
)

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

Streaming with OpenAI SDK¶

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:8000/openai/v1",
    api_key="dummy-key"
)

stream = client.chat.completions.create(
    model="claude-3-5-sonnet-20241022",
    messages=[{"role": "user", "content": "Tell me a story"}],
    stream=True,
    max_tokens=500
)

for chunk in stream:
    if chunk.choices[0].delta.content is not None:
        print(chunk.choices[0].delta.content, end="")

JavaScript/Node.js¶

import OpenAI from 'openai';

const openai = new OpenAI({
  baseURL: 'http://localhost:8000/openai/v1',
  apiKey: 'dummy-key' // Required but not used
});

const completion = await openai.chat.completions.create({
  model: 'claude-3-5-sonnet-20241022',
  messages: [{ role: 'user', content: 'Hello!' }],
  max_tokens: 100
});

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

curl¶

curl -X POST http://localhost:8000/openai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer dummy-key" \
  -d '{
    "model": "claude-3-5-sonnet-20241022",
    "messages": [{"role": "user", "content": "Hello!"}],
    "max_tokens": 100
  }'

Migration from OpenAI¶

To migrate existing OpenAI-based applications:

Change the base URL:

# Before
client = OpenAI(api_key="your-openai-key")

# After
client = OpenAI(
    base_url="http://localhost:8000/openai/v1",
    api_key="dummy-key"
)

Update model names to Claude models:

# Before
model="gpt-4-turbo"

# After
model="claude-3-5-sonnet-20241022"

Remove unsupported parameters (optional):
presence_penalty
frequency_penalty
user

That's it! Your existing OpenAI code should work without other changes.