[BUG] failed while using graphiti-core-0.24.x Integrated with Zhipu AI Model(GLM-4.6)

[sparkychen]

Bug Description

hi all,
I got an error during using graphiti-core-0.24.x via Zhipu AI model(GLM-4.6~.4.7), but I got well done when running the same code using aliyun qwen3 series models.
My code as below:

uv add -u langchain
uv add graphiti-core

import os
from config import settings
from langchain_openai import ChatOpenAI, OpenAIEmbeddings
from graphiti_core.driver.neo4j_driver import Neo4jDriver
from graphiti_core import Graphiti
from graphiti_core.llm_client.config import LLMConfig
from graphiti_core.llm_client.openai_generic_client import OpenAIGenericClient
from graphiti_core.embedder.openai import OpenAIEmbedder, OpenAIEmbedderConfig
from graphiti_core.cross_encoder.openai_reranker_client import OpenAIRerankerClient, CrossEncoderClient
from graphiti_core.cross_encoder.bge_reranker_client import BGERerankerClient
from openai import AsyncOpenAI
from dotenv import load_dotenv

load_dotenv()

ZHIPUAI_BASE_URL = os.getenv("ZHIPUAI_BASE_URL")
ZHIPUAI_API_KEY = os.getenv("ZHIPUAI_API_KEY")
ZHIPUAI_MODEL = os.getenv("ZHIPUAI_MODEL")
ZHIPUAI_EMBEDDING_MODEL = os.getenv("ZHIPUAI_EMBEDDING_MODEL")
ZHIPUAI_RERANK_MODEL = os.getenv("ZHIPUAI_RERANK_MODEL")
ZHIPUAI_TEMPERATURE = os.getenv("ZHIPUAI_TEMPERATURE")


neo4j_driver = Neo4jDriver(
    uri=settings.NEO4J_URI, 
    user=settings.NEO4J_USER, 
    password=settings.NEO4J_PASSWORD, 
    database=settings.NEO4J_DATABASE,
)

llm_config = LLMConfig(
    api_key=ZHIPUAI_API_KEY,
    model=ZHIPUAI_MODEL,
    # small_model=LLM_MODEL,
    base_url=ZHIPUAI_BASE_URL,
    max_tokens=8192,
    temperature=0.75
)
llm_client = OpenAIGenericClient(config=llm_config)
embedder_config = OpenAIEmbedderConfig(api_key=ZHIPUAI_API_KEY, base_url=ZHIPUAI_BASE_URL, embedding_model=ZHIPUAI_EMBEDDING_MODEL, embedding_dim=1024)
embedder = OpenAIEmbedder(config=embedder_config)
cross_encoder = OpenAIRerankerClient(
    config=LLMConfig(
        api_key=ZHIPUAI_API_KEY,
        model=ZHIPUAI_RERANK_MODEL,
        base_url=ZHIPUAI_BASE_URL), 
    client=llm_client,
)

graphiti = Graphiti(
    graph_driver=neo4j_driver,
    llm_client=llm_client,
    embedder=embedder,
    cross_encoder=cross_encoder,
    max_coroutines=20
)

try:
    resp1 = await graphiti.add_episode(
        name="用户咨询记录",
        episode_body="王强在2025-10-26日来电，表示对Xiaomi 17 pro max的续航能力、散热、音质和摄影功能非常感兴趣，他的预算是5000-7000元。",
        source=EpisodeType.text,
        source_description="客户服务对话记录",
        reference_time=datetime.now(timezone.utc)
    )
    # print(f"成功添加片段: {resp1}")
except EntityTypeValidationError as e:
    print(f"数据验证失败: {e.message}")
    traceback.print_exc()

Environment

• Graphiti Version: 0.24.3
• Python Version:3.11.14
• Operating System:Ubuntu 24.04
• Database Backend: Neo4j 5.26
• LLM Provider & Model: Zhipu AI GLM-4.6

Installation Method

• pip install
• [ *] uv add
• Development installation (git clone)

Error Messages/Traceback

error message like this:
2025-12-23 21:43:18 - graphiti_core.llm_client.openai_generic_client - ERROR - Error in generating LLM response: Expecting value: line 1 column 1 (char 0)
2025-12-23 21:43:18 - graphiti_core.llm_client.openai_generic_client - WARNING - Retrying after application error (attempt 1/2): Expecting value: line 1 column 1 (char 0)
2025-12-23 21:43:35 - graphiti_core.llm_client.openai_generic_client - ERROR - Error in generating LLM response: Expecting value: line 1 column 1 (char 0)
2025-12-23 21:43:35 - graphiti_core.llm_client.openai_generic_client - WARNING - Retrying after application error (attempt 2/2): Expecting value: line 1 column 1 (char 0)
2025-12-23 21:43:50 - graphiti_core.llm_client.openai_generic_client - ERROR - Error in generating LLM response: Expecting value: line 1 column 1 (char 0)
2025-12-23 21:43:50 - graphiti_core.llm_client.openai_generic_client - ERROR - Max retries (2) exceeded. Last error: Expecting value: line 1 column 1 (char 0)
处理过程中出现错误: Expecting value: line 1 column 1 (char 0)

Configuration

# Relevant configuration or initialization code

Additional Context

• Does this happen consistently or intermittently?
• Which component are you using? (core library, REST server, MCP server)
• Any recent changes to your environment?
• Related issues or similar problems you've encountered?

Possible Solution

If you have ideas about what might be causing the issue or how to fix it, please share them here.

[abhaybhargav]

+1 on this GLM is an important model and it would be useful to have support for it