Framework Integrations

TokenSense provides native callback handlers for both LangChain and LlamaIndex. Instead of wrapping your LLM client with observe(), you simply attach the respective TokenSense callback handler to your LLM configuration. This provides full observability into complex agentic workflows, RAG pipelines, and multi-step reasoning chains.

LangChain

Import TokenSenseCallbackHandler and pass it into the callbacks list when invoking your LangChain models.

from tokensense import TokenSenseCallbackHandler
from langchain_groq import ChatGroq

llm = ChatGroq(model="llama-3.1-8b-instant")

# Attach the callback handler to track this invocation
response = llm.invoke(
    "Hello!", 
    config={"callbacks": [TokenSenseCallbackHandler()]}
)

LlamaIndex

Import TokenSenseLlamaIndexCallback and add it to your global CallbackManager before running your query engines or agents.

from tokensense import TokenSenseLlamaIndexCallback
from llama_index.core.callbacks import CallbackManager
from llama_index.core import Settings
from llama_index.llms.groq import Groq

# Configure the global callback manager
Settings.callback_manager = CallbackManager([TokenSenseLlamaIndexCallback()])

llm = Groq(model="llama-3.1-8b-instant")
response = llm.complete("Hello!")