Chat and streaming

A chat completion

reply = client.chat.completions.create(
    model="fta/zai/glm-5.3",
    messages=[{"role": "user", "content": "Name three uses for a paperclip."}],
    temperature=0.7,
    max_tokens=400,
)
message = reply["choices"][0]["message"]
print(message["content"])

Any extra keyword argument is sent as is, so every Chat Completions parameter works: temperature, top_p, max_tokens, stop, seed, tools, tool_choice, response_format, reasoning_effort, web_search_options.

max_tokens=0

Sending max_tokens=0 asks for the longest reply the model allows.

Streaming text

for text in client.chat.completions.stream_text(
    model="fta/kimi/k3",
    messages=[{"role": "user", "content": "Tell me a short story."}],
):
    print(text, end="", flush=True)

Streaming chunks

stream=True returns an iterator of chat.completion.chunk dictionaries. The last chunk carries usage.

for chunk in client.chat.completions.create(model="fta/kimi/k3", messages=messages, stream=True):
    for choice in chunk["choices"]:
        print(choice["delta"].get("content") or "", end="")
    if chunk.get("usage"):
        print("\n", chunk["usage"])

If the stream stops before the model finishes, the iterator raises StreamInterruptedError with the text received so far in .partial.

Reasoning

Thinking models return their reasoning in message["reasoning_content"] (and delta["reasoning_content"] when streaming). Turn thinking off with reasoning_effort="none":

reply = client.chat.completions.create(model="fta/zai/glm-5.3", messages=messages, reasoning_effort="none")

Some models write their thinking at the start of the reply inside <think>...</think>; the client moves it to reasoning_content so content holds only the answer.

Images in messages

Models that read images take OpenAI-style content parts:

reply = client.chat.completions.create(
    model="fta/zai/glm-5.3-flash",
    messages=[{
        "role": "user",
        "content": [
            {"type": "text", "text": "What is in this picture?"},
            {"type": "image_url", "image_url": {"url": "https://example.com/cat.jpg"}},
        ],
    }],
)

A text-only model answers with ImagesNotSupportedError; with fallback models set, the next model is tried.

Which model answered

With fallback models, reply["model"] names the model that answered.