llm-api-adapter-zai

An optional Z.ai / GLM organization package for LLM API Adapter, a Python SDK with one shared interface for calling LLM APIs.

Uses Z.ai / GLM's official Chat Completions API without a provider SDK.

Installation

Install through the Core package extra:

pip install "llm-api-adapter[zai]"

Direct installation is also supported when Core is managed separately:

pip install llm-api-adapter-zai

Async requests need HTTPX:

pip install "llm-api-adapter[zai,async]"

Synchronous requests use requests by default. To opt into HTTPX for sync chat() and stream_chat(), install "llm-api-adapter[zai,httpx]" and pass transport="httpx".

Quick start

import os

from llm_api_adapter.models.messages.chat_message import UserMessage
from llm_api_adapter.universal_adapter import UniversalLLMAPIAdapter

adapter = UniversalLLMAPIAdapter(
    organization="zai",
    model="glm-5.3-flash",
    api_key=os.environ["ZAI_API_KEY"],
)

response = adapter.chat(
    messages=[UserMessage("Explain retrieval-augmented generation.")],
    max_tokens=128,
)
print(response.content)

Supported models

• glm-5.3-flash
• glm-5.3-flashx

Capabilities

Unknown, retired, aliased, or unverified model IDs are not inferred. Unsupported tool choices, parallel-tool control, provider-built-in tools, and every DocumentPart URL or byte form fail locally before the transport is called. The adapter accepts only the official endpoint and the ZAI_API_KEY bearer credential.

reasoning_level="none" uses the minimum registered effort and emits a warning. previous_response is accepted but ignored: send complete messages history, including assistant tool calls and tool results, on each turn; no provider continuation identifier is sent. Exact values and exceptions are recorded in the Z.ai registry.

Automatic cache usage and pricing

usage.prompt_tokens_details.cached_tokens becomes Usage.cached_tokens, a subset of total input. The model uses its registry's ordinary and cache-read USD rates. No separately priced automatic cache-write component is registered, so cache_write_tokens remains None.

An omitted cache-read split leaves cost_input and cost_total unknown; known output cost can remain available. The adapter does not enable opt-in cache controls or price selectable TTLs or storage. See the shared usage and pricing guide.

See the main llm-api-adapter README for the shared API contract and transport behavior.