Skip to main content
SparkLLM chat models API by iFlyTek. For more information, see iFlyTek Open Platform.

​
Basic use

"""For basic init and call"""
from langchain_community.chat_models import ChatSparkLLM
from langchain.messages import HumanMessage

chat = ChatSparkLLM(
    spark_app_id="<app_id>", spark_api_key="<api_key>", spark_api_secret="<api_secret>"
)
message = HumanMessage(content="Hello")
chat([message])

AIMessage(content='Hello! How can I help you today?')
  • Get SparkLLM’s app_id, api_key and api_secret from iFlyTek SparkLLM API Console (for more info, see iFlyTek SparkLLM Intro ), then set environment variables IFLYTEK_SPARK_APP_ID, IFLYTEK_SPARK_API_KEY and IFLYTEK_SPARK_API_SECRET or pass parameters when creating ChatSparkLLM as the demo above.

​
For ChatSparkLLM with Streaming

chat = ChatSparkLLM(
    spark_app_id="<app_id>",
    spark_api_key="<api_key>",
    spark_api_secret="<api_secret>",
    streaming=True,
)
for chunk in chat.stream("Hello!"):
    print(chunk.content, end="")

Hello! How can I help you today?

​
For v2

"""For basic init and call"""
from langchain_community.chat_models import ChatSparkLLM
from langchain.messages import HumanMessage

chat = ChatSparkLLM(
    spark_app_id="<app_id>",
    spark_api_key="<api_key>",
    spark_api_secret="<api_secret>",
    spark_api_url="wss://spark-api.xf-yun.com/v2.1/chat",
    spark_llm_domain="generalv2",
)
message = HumanMessage(content="Hello")
chat([message])
Edit the source of this page on GitHub.
Connect these docs programmatically to Claude, VSCode, and more via MCP for real-time answers.