Z.ai (GLM)¶

Use the zai provider to call GLM models with an API key. Chat uses Chat Completions.

Get started¶

Install Republic using the quickstart, set REPUBLIC_ZAI_API_KEY, and replace MODEL_ID with a model available to your account:

import asyncio

import republic


async def main():
    model = republic.get_model("zai:MODEL_ID")
    response = await model.chat("Say hello in one sentence.")
    print(response.text)


asyncio.run(main())

Authentication¶

Republic reads REPUBLIC_ZAI_API_KEY when the provider is created and sends the key in the Authorization: Bearer header. You can pass a key loaded by your application as api_key=; see API-key authentication.

Model support¶

The default endpoint is https://api.z.ai/api/paas/v4; set api_base="https://open.bigmodel.cn/api/paas/v4" for the China platform. Chat requests send max_tokens. Thinking models return reasoning in response.reasoning. Republic sends reasoning_content back with earlier assistant messages, which these models require across tool calls; keep response.message in the conversation. Z.ai accepts only tool_choice="auto".

See configuration for endpoint overrides and request options.