OpenAI SDK compatibility

OxyRouter implements the OpenAI chat completions API. The official SDKs work by changing the base URL and API key.

Python

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["OXYROUTER_API_KEY"],
    base_url="https://api.oxyrouter.in/v1",
)

completion = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "Hello"}],
)

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

Node.js

import OpenAI from 'openai'

const client = new OpenAI({
  apiKey: process.env.OXYROUTER_API_KEY,
  baseURL: 'https://api.oxyrouter.in/v1',
})

const completion = await client.chat.completions.create({
  model: 'gpt-4o-mini',
  messages: [{ role: 'user', content: 'Hello' }],
})

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

Streaming

Python

stream = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "Count to five."}],
    stream=True,
)

for chunk in stream:
    delta = chunk.choices[0].delta.content
    if delta:
        print(delta, end="", flush=True)

Node.js

const stream = await client.chat.completions.create({
  model: 'gpt-4o-mini',
  messages: [{ role: 'user', content: 'Count to five.' }],
  stream: true,
})

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content ?? '')
}

Tool calling

Tools are declared exactly as in the OpenAI API. Send the assistant’s tool_calls back followed by one tool message per result.

tools = [
    {
        "type": "function",
        "function": {
            "name": "get_weather",
            "description": "Get current weather for a city",
            "parameters": {
                "type": "object",
                "properties": {"city": {"type": "string"}},
                "required": ["city"],
            },
        },
    }
]

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "Weather in Pune?"}],
    tools=tools,
)

tool_call = response.choices[0].message.tool_calls[0]

Error handling

Errors use the same envelope as OpenAI. The SDK raises an error with the HTTP status, code, and message.

from openai import APIStatusError

try:
    completion = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": "Hello"}],
    )
except APIStatusError as error:
    print(error.status_code)
    print(error.body)

Common cases:

Status	Meaning
401	Invalid or expired API key
402	Wallet balance or spending cap prevents the request
403	Missing scope or model not allowed
404	Unknown or unavailable model
429	Rate or concurrency limit
502 / 503	Upstream or platform failure

See Errors and limits for codes and retry guidance.

Configuration tips

• Set the SDK’s base URL to https://api.oxyrouter.in/v1 and keep the default paths.
• Pick model IDs from GET /v1/models. Do not reuse another provider’s internal model names.
• Keep max_tokens or max_completion_tokens set, but never both. A cap also makes spending predictable.
• Disable SDK-level request retries for streaming, or set a single retry with a timeout; a stream cannot be replayed mid-flight.
• For non-streaming requests, add an Idempotency-Key header when your SDK supports custom headers, so retries do not double-charge.
• Read X-Request-ID from responses and log it for support.

Unsupported OpenAI endpoints

This reference covers chat completions and model discovery. Responses, embeddings, audio, images, and other OpenAI endpoints are not part of the documented OxyRouter API surface. Calling an undocumented path may return 404.