API Documentation

Set stream: true on /v1/chat/completions to receive incremental tokens. Each SSE frame is a JSON delta; the last frame before [DONE] contains the final usage block when stream_options.include_usage: true.

# Python
from openai import OpenAI
client = OpenAI(api_key="sk-svp-...", base_url="https://api.genon.ai/v1")
stream = client.chat.completions.create(
    model="zai-org/glm-5.2",
    messages=[{"role": "user", "content": "Tell me a story"}],
    stream=True,
    stream_options={"include_usage": True},
)
for chunk in stream:
    delta = chunk.choices[0].delta.content if chunk.choices else None
    if delta:
        print(delta, end="", flush=True)
    if chunk.usage:
        print(f"\ntokens: in={chunk.usage.prompt_tokens} out={chunk.usage.completion_tokens}")

All chat models advertise the tools feature. Pass an OpenAI-style tools array and tool_choice ("auto" / "required" / {type:"function",function:{name}}). The selected model picks tools and emits tool_calls[] in choices[0].message; reply with a role:"tool" message to feed execution results back.

r = client.chat.completions.create(
    model="zai-org/glm-5.2",
    messages=[{"role": "user", "content": "What's the weather in Seoul?"}],
    tools=[{
        "type": "function",
        "function": {
            "name": "get_weather",
            "description": "Get current weather for a location",
            "parameters": {
                "type": "object",
                "properties": {"location": {"type": "string"}},
                "required": ["location"],
            },
        },
    }],
    tool_choice="auto",
)
call = r.choices[0].message.tool_calls[0]
print(call.function.name, call.function.arguments)

GLM-5.2, Qwen3.5, and Kimi-K2.6 emit chain-of-thought tokens separately as message.reasoning_content (and mirrored under provider_specific_fields.reasoning). They are billed as completion tokens. Toggle thinking off via chat_template_kwargs when the model accepts it (Kimi-K2.6 supported, qwen3.5 ignores the flag).

r = client.chat.completions.create(
    model="moonshotai/kimi-k2.6",
    messages=[{"role": "user", "content": "Explain prime factorisation"}],
    extra_body={"chat_template_kwargs": {"thinking": False}},
)
msg = r.choices[0].message
print("answer:", msg.content)
print("thinking:", msg.reasoning_content)  # may be None when disabled

qwen/qwen3.5-397b-a17b-fp8 accepts text + image input. Send the OpenAI multimodal content array; image URLs may be http(s):// or data:image/...;base64,....

r = client.chat.completions.create(
    model="qwen/qwen3.5-397b-a17b-fp8",
    messages=[{
        "role": "user",
        "content": [
            {"type": "text", "text": "What's in this image?"},
            {"type": "image_url", "image_url": {"url": "https://example.com/cat.png"}},
        ],
    }],
)
print(r.choices[0].message.content)

The serving API is OpenAI-compatible — point any OpenAI SDK at our base URL.

from langchain_openai import ChatOpenAI
llm = ChatOpenAI(
    model="zai-org/glm-5.2",
    openai_api_key="sk-svp-...",
    openai_api_base="https://api.genon.ai/v1",
)
print(llm.invoke("hi"))

{
  "models": [{
    "title": "GenOn GLM-5.2",
    "provider": "openai",
    "model": "zai-org/glm-5.2",
    "apiBase": "https://api.genon.ai/v1",
    "apiKey": "sk-svp-..."
  }]
}

OPENAI_API_BASE=https://api.genon.ai/v1 \
OPENAI_API_KEY=sk-svp-... \
aider --model openai/zai-org/glm-5.2

# Both /v1/messages and /v1/chat/completions work — see the Chat tab above
# for the Anthropic shape (#123/#147).