API Reference

#Access methods

Rungate supports three ways to access the inference API: account-scoped API keys, x402 pay-per-request, and MPP pay-per-request. Environments can advertise one or both payment protocols when a request arrives without standard authentication.

#Configuration

To start using Rungate with OpenAI's client libraries, pass your Rungate API key and change the SDK base URL to https://api.rungate.ai/v1. For raw HTTP requests, use https://api.rungate.ai with /v1/... paths. You can find your API key in your account settings.

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_RUNGATE_API_KEY",
    base_url="https://api.rungate.ai/v1",
)

#Chat completions

Once your client is configured, you can query any of our open-source models. For example, here's a chat completion with DeepSeek V3.2.

response = client.chat.completions.create(
    model="deepseek/deepseek-v3.2",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "What is the capital of France?"},
    ],
)

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

See the Models & Pricing page for all available model IDs.

#Streaming

You can stream responses back using OpenAI's streaming interface. Pass stream=True to receive server-sent events as the model generates tokens.

stream = client.chat.completions.create(
    model="qwen/qwen3-coder-next",
    messages=[{"role": "user", "content": "Count to 5"}],
    stream=True,
)

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

The final chunk before [DONE] includes a usage field with token counts.

#Multimodal input

Models that support vision accept image inputs alongside text. Pass a content array instead of a plain string.

response = client.chat.completions.create(
    model="qwen/qwen3-vl-235b-a22b-thinking",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "What is in this image?"},
                {
                    "type": "image_url",
                    "image_url": {
                        "url": "https://example.com/photo.jpg"
                    },
                },
            ],
        }
    ],
)

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

Supported formats: image/jpeg, image/png, image/webp, image/gif. Base64 data URIs and public URLs are both accepted. See Models & Pricing for which models support vision.

#Endpoints

These paths are also available without the /v1 prefix (e.g. /chat/completions, /models).

#Usage tracking

Every response includes a usage object with token counts. For streaming responses it appears in the final chunk before [DONE].

{
  "usage": {
    "prompt_tokens": 12,
    "completion_tokens": 8,
    "total_tokens": 20
  }
}

#Rate limiting

Throughput limits vary per model — see the Models & Pricing page for details.

#Error codes

Errors follow the standard OpenAI error format:

{
  "error": {
    "code": 401,
    "message": "Invalid API key"
  }
}