API quick start — curl, Python, TypeScript

The fastest way to test the API is curl:

Normal request:

curl https://gate.joingonka.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer jg-your-key" \
  -d '{
    "model": "Qwen/Qwen3-235B-A22B-Instruct-2507-FP8",
    "messages": [
      {"role": "user", "content": "What is Gonka?"}
    ]
  }'

Streaming (response in parts):

curl https://gate.joingonka.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer jg-your-key" \
  -d '{
    "model": "Qwen/Qwen3-235B-A22B-Instruct-2507-FP8",
    "messages": [
      {"role": "user", "content": "Write hello world in Python"}
    ],
    "stream": true
  }'

The response comes in JSON format (normal) or Server-Sent Events (streaming) — fully compatible with the OpenAI API.

The official OpenAI Python SDK works with JoinGonka Gateway without changes:

pip install openai

Normal request:

from openai import OpenAI

client = OpenAI(
    base_url="https://gate.joingonka.ai/v1",
    api_key="jg-your-key",
)

response = client.chat.completions.create(
    model="Qwen/Qwen3-235B-A22B-Instruct-2507-FP8",
    messages=[
        {"role": "user", "content": "Explain blockchain in simple terms"}
    ],
)

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

Streaming:

stream = client.chat.completions.create(
    model="Qwen/Qwen3-235B-A22B-Instruct-2507-FP8",
    messages=[{"role": "user", "content": "Write sorting in Python"}],
    stream=True,
)

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

Tool calling:

import json

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

response = client.chat.completions.create(
    model="Qwen/Qwen3-235B-A22B-Instruct-2507-FP8",
    messages=[{"role": "user", "content": "What is the weather in Moscow?"}],
    tools=tools,
)

tool_call = response.choices[0].message.tool_calls[0]
print(f"Function: {tool_call.function.name}")
print(f"Arguments: {tool_call.function.arguments}")

Qwen3-235B supports native tool calling — functions are called correctly, without parsing text responses.

Installation:

npm install openai

Normal request:

import OpenAI from 'openai';

const client = new OpenAI({
  baseURL: 'https://gate.joingonka.ai/v1',
  apiKey: 'jg-your-key',
});

async function main() {
  const response = await client.chat.completions.create({
    model: 'Qwen/Qwen3-235B-A22B-Instruct-2507-FP8',
    messages: [
      { role: 'user', content: 'Write an Express.js server' },
    ],
  });

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

main();

Streaming:

const stream = await client.chat.completions.create({
  model: 'Qwen/Qwen3-235B-A22B-Instruct-2507-FP8',
  messages: [{ role: 'user', content: 'Explain async/await' }],
  stream: true,
});

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

Tool calling:

const response = await client.chat.completions.create({
  model: 'Qwen/Qwen3-235B-A22B-Instruct-2507-FP8',
  messages: [{ role: 'user', content: 'Convert 100 USD to EUR' }],
  tools: [{
    type: 'function',
    function: {
      name: 'convert_currency',
      description: 'Currency conversion',
      parameters: {
        type: 'object',
        properties: {
          amount: { type: 'number' },
          from: { type: 'string' },
          to: { type: 'string' },
        },
        required: ['amount', 'from', 'to'],
      },
    },
  }],
});

const toolCall = response.choices[0].message.tool_calls?.[0];
console.log(`Function: ${toolCall?.function.name}`);
console.log(`Arguments: ${toolCall?.function.arguments}`);

All examples use the official OpenAI SDK — no additional libraries are required. Simply replace base_url and api_key.

JoinGonka Gateway supports all standard OpenAI Chat Completions API parameters:

Qwen3-235B model parameters: context window — 128K tokens, maximum response — up to 8192 tokens (Kimi K2.6 — 3072, MiniMax M2.7 — 4096). Full specifications: HuggingFace. The list of models is available via GET /v1/models.

Two endpoints:

• OpenAI-format: POST https://gate.joingonka.ai/v1/chat/completions
• Anthropic-format: POST https://gate.joingonka.ai/v1/messages

Authentication: Authorization: Bearer jg-your-key (OpenAI) or x-api-key: jg-your-key (Anthropic)

The response format is fully compatible with OpenAI and Anthropic — any SDK, library, or framework supporting OpenAI or Anthropic API works with JoinGonka Gateway without modifications. Claude Code connects directly via Anthropic format.