DeepAnima
切换到中文
Kouri Ai

OpenAI Compatible API

Use OpenAI SDK to call various models on Kouri Ai

Kouri Ai is fully compatible with OpenAI API format. You can use the official OpenAI SDK directly by simply changing the base_url and api_key to call all supported models.

Endpoints

Endpoint TypeURLDescription
Chat Completionshttps://api.kourichat.com/v1/chat/completionsChat completion API
Responseshttps://api.kourichat.com/v1/responsesResponse API
Standard Endpointhttps://api.kourichat.com/v1Recommended for SDKs (with /v1)
Base Endpointhttps://api.kourichat.comFor some applications

When configuring base_url, make sure to add the /v1 suffix, otherwise you may get 404 errors.

API Selection Guide

OpenAI provides two main chat APIs. Please choose based on model requirements:

APIPathSupported ModelsMax Timeout
Chat Completions/v1/chat/completionsGPT-4o, GPT-4, most models~5 min
Responses/v1/responsesgpt-5.2-pro, o3-pro, etc.~20 min

Chat Completions API

Suitable for most scenarios, supports regular chat models:

POST /v1/chat/completions

Response API (Required for Reasoning Models)

Important: Some advanced models like gpt-5.2-pro and o3-pro only support the Response API, not Chat Completions.

Advantages of Response API:

  • Longer timeout: Up to ~20 minutes, suitable for complex reasoning
  • Better reasoning support: Designed for o-series reasoning models
  • More advanced features: Such as reasoning token budget
POST /v1/responses

For models that support both APIs (like o1, o3), Kouri Ai handles compatibility automatically. You can use Chat Completions and the platform will convert. But for models that only support Response API, you must use the Response API.

Sora Video Generation API

This API is special. Please refer to the official documentation or Apifox for usage.

Quick Start

cURL Request

Chat Completions API:

curl https://api.kourichat.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-xxxxxxxx" \
  -d '{
    "model": "gpt-4o",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Response API (required for gpt-5.2-pro, etc.):

curl https://api.kourichat.com/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-xxxxxxxx" \
  -d '{
    "model": "gpt-5.2-pro",
    "input": "Explain the basic principles of quantum computing"
  }'

Python

Using the official OpenAI Python SDK:

Chat Completions API:

from openai import OpenAI

client = OpenAI(
    base_url='https://api.kourichat.com/v1',
    api_key='sk-xxxxxxxx',  # Replace with your Kouri Ai token
)

# Standard request
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "user", "content": "Hello!"}
    ]
)

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

Response API (for gpt-5.2-pro, etc.):

from openai import OpenAI

client = OpenAI(
    base_url='https://api.kourichat.com/v1',
    api_key='sk-xxxxxxxx',
)

# Use Response API for gpt-5.2-pro
response = client.responses.create(
    model="gpt-5.2-pro",
    input="Explain the basic principles of quantum computing"
)

print(response.output_text)

Response API - with reasoning parameters:

from openai import OpenAI

client = OpenAI(
    base_url='https://api.kourichat.com/v1',
    api_key='sk-xxxxxxxx',
)

response = client.responses.create(
    model="gpt-5.2-pro",
    input="How do I solve this math problem?",
    reasoning={
        "effort": "high"  # Reasoning depth: low, medium, high
    }
)

print(response.output_text)

Response API - Streaming:

from openai import OpenAI

client = OpenAI(
    base_url='https://api.kourichat.com/v1',
    api_key='sk-xxxxxxxx',
)

# Streaming with Response API
stream = client.responses.create(
    model="gpt-5.2-pro",
    input="Tell me a story",
    stream=True
)

for event in stream:
    if hasattr(event, 'type') and event.type == 'response.output_text.delta':
        print(event.delta, end="", flush=True)

Streaming

from openai import OpenAI

client = OpenAI(
    base_url='https://api.kourichat.com/v1',
    api_key='sk-xxxxxxxx',
)

# Streaming request
stream = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "user", "content": "Tell me a story"}
    ],
    stream=True
)

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

JavaScript / Node.js

Using the official OpenAI Node.js SDK:

import OpenAI from 'openai';

const client = new OpenAI({
    baseURL: 'https://api.kourichat.com/v1',
    apiKey: 'sk-xxxxxxxx',  // Replace with your Kouri Ai token
});

async function main() {
    const response = await client.chat.completions.create({
        model: 'gpt-4o',
        messages: [
            { role: 'user', content: 'Hello!' }
        ]
    });

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

main();

Streaming

import OpenAI from 'openai';

const client = new OpenAI({
    baseURL: 'https://api.kourichat.com/v1',
    apiKey: 'sk-xxxxxxxx',
});

async function main() {
    const stream = await client.chat.completions.create({
        model: 'gpt-4o',
        messages: [{ role: 'user', content: 'Tell me a story' }],
        stream: true,
    });

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

main();

LangChain Integration

Configure via environment variables:

import os

os.environ["OPENAI_API_BASE"] = "https://api.kourichat.com/v1"
os.environ["OPENAI_API_KEY"] = "sk-xxxxxxxx"

Or configure directly in code:

from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    model="gpt-4o",
    base_url="https://api.kourichat.com/v1",
    api_key="sk-xxxxxxxx",
)

response = llm.invoke("Hello!")
print(response.content)

LangChain's OPENAI_API_BASE environment variable requires the /v1 suffix.

Multimodal Models

Image Understanding

Supports image URLs or Base64-encoded images:

from openai import OpenAI

client = OpenAI(
    base_url='https://api.kourichat.com/v1',
    api_key='sk-xxxxxxxx',
)

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "What's in this image?"},
                {
                    "type": "image_url",
                    "image_url": {
                        "url": "https://example.com/image.jpg"
                    }
                }
            ]
        }
    ]
)

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

Using Base64 Images

import base64
from openai import OpenAI

client = OpenAI(
    base_url='https://api.kourichat.com/v1',
    api_key='sk-xxxxxxxx',
)

# Read and encode local image
with open("image.jpg", "rb") as image_file:
    base64_image = base64.b64encode(image_file.read()).decode('utf-8')

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "Describe this image"},
                {
                    "type": "image_url",
                    "image_url": {
                        "url": f"data:image/jpeg;base64,{base64_image}"
                    }
                }
            ]
        }
    ]
)

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

Common Issues

404 Error

Make sure base_url has the /v1 suffix:

# Correct
base_url='https://api.kourichat.com/v1'

# Wrong
base_url='https://api.kourichat.com'

Legacy SDK Compatibility

If using an older OpenAI SDK (< 1.0), configuration is slightly different:

import openai

openai.api_base = "https://api.kourichat.com/v1"
openai.api_key = "sk-xxxxxxxx"

We recommend upgrading to OpenAI SDK 1.0+ for a better experience.

More Resources

API Overview

Kouri Ai API basics and usage guide

Anthropic Compatible API

Call Claude models using native Anthropic protocol or OpenAI-compatible protocol

On this page

EndpointsAPI Selection GuideChat Completions APIResponse API (Required for Reasoning Models)Sora Video Generation APIQuick StartcURL RequestPythonStreamingJavaScript / Node.jsStreamingLangChain IntegrationMultimodal ModelsImage UnderstandingUsing Base64 ImagesCommon Issues404 ErrorLegacy SDK CompatibilityMore Resources