Tool Calling Code

A developer published a walkthrough for implementing tool calling in a LangChain agent, using the @tool decorator to expose Python functions such as get_weather and add_numbers, binding them to a Groq-hosted model via bind_tools, and manually executing the model's tool_calls before a second LLM invocation produces the final answer. The example also covers virtual environment setup and a pinned requirements.txt that includes langchain, langchain-groq, langgraph, and chromadb.

python -m venv .venv python3 -m venv .venv apt install python3.12-venv python3.12 -m venv .venv source .venv/bin/activate python3 <python file pip install -r requirements.txt If virtual environment to be removed, rm -rf .venv user , first reply and then agent. python from dotenv import load dotenv from langchain core.messages import HumanMessage from langchain core.tools import tool from langchain groq import ChatGroq load dotenv @tool Decorator def get weather city: str - str: """ Get the current weather for a given city. city: name of the city. """ Docstrings Replace with a real weather API call return f"The weather in {city} is sunny and 25°C." @tool def add numbers a: int, b: int - int: """Add two numbers together.""" return a + b tools = get weather, add numbers tool map = {t.name: t for t in tools} print tool map llm = ChatGroq model="openai/gpt-oss-120b", temperature=0 llm with tools = llm.bind tools tools def run query question: str - str: messages = HumanMessage question print "Initial Messages ", messages input "Wait ...." First call: model decides whether to answer directly or call a tool ai msg = llm with tools.invoke messages messages.append ai msg print "Messages after first call ", messages input "Wait ...." if ai msg.tool calls: print "tool calls ", ai msg.tool calls input "wait ..." Execute each requested tool call for call in ai msg.tool calls: selected tool = tool map call "name" tool result = selected tool.invoke call "args" messages.append { "role": "tool", "content": str tool result , "tool call id": call "id" , } print "Messages after tool call ", messages input "Wait ...." Second call: let the model turn tool results into a final answer print "Total Messages ", messages input "Wait Final...." final response = llm with tools.invoke messages return final response.content else: No tool needed — first response is already the final answer return ai msg.content if name == " main ": print run query "Give me climate in paris and london?" print run query "What is 15 plus 27?" print run query "convert 100rs to usd" aiohappyeyeballs==2.7.1 aiohttp==3.14.3 aiosignal==1.4.0 annotated-doc==0.0.5 annotated-types==0.8.0 anyio==4.14.2 asgiref==3.12.1 async-timeout==4.0.3 attrs==26.1.0 bcrypt==5.0.0 build==1.6.1 certifi==2026.7.22 charset-normalizer==3.5.1 chromadb==1.5.9 click==8.5.0 coloredlogs==15.0.1 distro==1.9.0 durationpy==0.11 exceptiongroup==1.3.1 fastapi==0.141.1 filelock==3.32.6 flatbuffers==25.12.19 frozenlist==1.8.0 fsspec==2026.7.0 googleapis-common-protos==1.75.3 greenlet==3.5.5 groq==0.37.1 grpcio==1.83.1 h11==0.16.0 hf-xet==1.6.0 httpcore==1.0.9 httpcore2==2.12.0 httptools==0.8.0 httpx==0.28.1 httpx-sse==0.4.3 httpx2==2.12.0 huggingface hub==1.31.0 humanfriendly==10.0 idna==3.19 importlib resources==7.1.0 jsonpatch==1.33 jsonpointer==3.1.1 jsonschema==4.26.0 jsonschema-specifications==2025.9.1 kubernetes==36.0.3 langchain==1.3.18 langchain-chroma==1.1.0 langchain-classic==1.0.8 langchain-community==0.4.2 langchain-core==1.6.1 langchain-groq==1.1.3 langchain-ollama==1.1.0 langchain-protocol==0.0.19 langchain-text-splitters==1.1.2 langgraph==1.2.11 langgraph-checkpoint==4.2.0 langgraph-prebuilt==1.1.0 langgraph-sdk==0.4.4 langsmith==0.12.1 markdown-it-py==4.2.0 mdurl==0.1.2 mmh3==5.3.0 mpmath==1.3.0 multidict==6.8.0 numpy==2.2.6 oauthlib==3.3.1 ollama==0.6.2 onnxruntime==1.23.2 opentelemetry-api==1.44.0 opentelemetry-exporter-otlp-proto-common==1.44.0 opentelemetry-exporter-otlp-proto-grpc==1.44.0 opentelemetry-proto==1.44.0 opentelemetry-sdk==1.44.0 opentelemetry-semantic-conventions==0.65b0 orjson==3.12.0 ormsgpack==1.12.2 overrides==7.7.0 packaging==26.3 propcache==0.5.2 protobuf==7.36.1 pybase64==1.5.0 pydantic==2.13.5 pydantic-settings==2.15.0 pydantic core==2.46.5 Pygments==2.21.0 pypdf==6.18.1 PyPika==0.51.1 pyproject hooks==1.2.0 python-dateutil==2.9.0.post0 python-dotenv==1.2.3 PyYAML==6.0.3 referencing==0.37.0 requests==2.34.2 requests-oauthlib==2.0.0 requests-toolbelt==1.0.0 rich==15.0.0 rpds-py==0.30.0 ruff==0.16.5 shellingham==1.5.4 six==1.17.0 sniffio==1.3.1 SQLAlchemy==2.0.52 sqlparse==0.6.0 starlette==1.6.0 sympy==1.14.0 tenacity==9.1.4 tokenizers==0.23.2 tomli==2.4.1 tqdm==4.70.1 truststore==0.10.4 typer==0.27.2 typing-inspection==0.4.4 typing extensions==4.16.0 urllib3==2.7.0 uuid utils==0.17.0 uvicorn==0.52.4 uvloop==0.22.1 watchfiles==1.2.0 websocket-client==1.9.2 websockets==16.1.1 xxhash==4.0.1 yarl==1.24.5 zstandard==0.25.0 GROQ API KEY="" python -m pipe freeze requirements.txt {'get weather': StructuredTool name='get weather', description='Get the current weather for a given city.\ncity: name of the city.', args schema=, func= , 'add numbers': StructuredTool name='add numbers', description='Add two numbers together.', args schema=, func= } Initial Messages HumanMessage content='Give me climate in Chennai and Mumbai?', additional kwargs={}, response metadata={} Messages after first call HumanMessage content='Give me climate in Chennai and Mumbai?', additional kwargs={}, response metadata={} , AIMessage content='', additional kwargs={'reasoning content': 'The user asks: "Give me climate in Chennai and Mumbai?" Likely they want current weather/climate. We have a function get weather that can get current weather for a city. We can call it for Chennai and Mumbai. Probably need to call twice. Use function calls.', 'tool calls': {'id': 'fc 8ee481c4-db27-4a4f-a4e8-839700ea29bb', 'function': {'arguments': '{"city":"Chennai"}', 'name': 'get weather'}, 'type': 'function'} }, response metadata={'token usage': {'completion tokens': 85, 'prompt tokens': 161, 'total tokens': 246, 'completion time': 0.176815886, 'completion tokens details': {'reasoning tokens': 57}, 'prompt time': 0.007317037, 'prompt tokens details': None, 'queue time': 0.34890584, 'total time': 0.184132923}, 'model name': 'openai/gpt-oss-120b', 'system fingerprint': 'fp 068241849b', 'service tier': 'on demand', 'finish reason': 'tool calls', 'logprobs': None, 'model provider': 'groq'}, id='lc run--01a0de65-11eb-75a2-b3b4-2770359461ad-0', tool calls= {'name': 'get weather', 'args': {'city': 'Chennai'}, 'id': 'fc 8ee481c4-db27-4a4f-a4e8-839700ea29bb', 'type': 'tool call'} , invalid tool calls= , usage metadata={'input tokens': 161, 'output tokens': 85, 'total tokens': 246, 'output token details': {'reasoning': 57}} tool calls {'name': 'get weather', 'args': {'city': 'Chennai'}, 'id': 'fc 8ee481c4-db27-4a4f-a4e8-839700ea29bb', 'type': 'tool call'} Messages after tool call HumanMessage content='Give me climate in Chennai and Mumbai?', additional kwargs={}, response metadata={} , AIMessage content='', additional kwargs={'reasoning content': 'The user asks: "Give me climate in Chennai and Mumbai?" Likely they want current weather/climate. We have a function get weather that can get current weather for a city. We can call it for Chennai and Mumbai. Probably need to call twice. Use function calls.', 'tool calls': {'id': 'fc 8ee481c4-db27-4a4f-a4e8-839700ea29bb', 'function': {'arguments': '{"city":"Chennai"}', 'name': 'get weather'}, 'type': 'function'} }, response metadata={'token usage': {'completion tokens': 85, 'prompt tokens': 161, 'total tokens': 246, 'completion time': 0.176815886, 'completion tokens details': {'reasoning tokens': 57}, 'prompt time': 0.007317037, 'prompt tokens details': None, 'queue time': 0.34890584, 'total time': 0.184132923}, 'model name': 'openai/gpt-oss-120b', 'system fingerprint': 'fp 068241849b', 'service tier': 'on demand', 'finish reason': 'tool calls', 'logprobs': None, 'model provider': 'groq'}, id='lc run--01a0de65-11eb-75a2-b3b4-2770359461ad-0', tool calls= {'name': 'get weather', 'args': {'city': 'Chennai'}, 'id': 'fc 8ee481c4-db27-4a4f-a4e8-839700ea29bb', 'type': 'tool call'} , invalid tool calls= , usage metadata={'input tokens': 161, 'output tokens': 85, 'total tokens': 246, 'output token details': {'reasoning': 57}} , {'role': 'tool', 'content': 'The weather in Chennai is sunny and 25°C.', 'tool call id': 'fc 8ee481c4-db27-4a4f-a4e8-839700ea29bb'} Total Messages HumanMessage content='Give me climate in Chennai and Mumbai?', additional kwargs={}, response metadata={} , AIMessage content='', additional kwargs={'reasoning content': 'The user asks: "Give me climate in Chennai and Mumbai?" Likely they want current weather/climate. We have a function get weather that can get current weather for a city. We can call it for Chennai and Mumbai. Probably need to call twice. Use function calls.', 'tool calls': {'id': 'fc 8ee481c4-db27-4a4f-a4e8-839700ea29bb', 'function': {'arguments': '{"city":"Chennai"}', 'name': 'get weather'}, 'type': 'function'} }, response metadata={'token usage': {'completion tokens': 85, 'prompt tokens': 161, 'total tokens': 246, 'completion time': 0.176815886, 'completion tokens details': {'reasoning tokens': 57}, 'prompt time': 0.007317037, 'prompt tokens details': None, 'queue time': 0.34890584, 'total time': 0.184132923}, 'model name': 'openai/gpt-oss-120b', 'system fingerprint': 'fp 068241849b', 'service tier': 'on demand', 'finish reason': 'tool calls', 'logprobs': None, 'model provider': 'groq'}, id='lc run--01a0de65-11eb-75a2-b3b4-2770359461ad-0', tool calls= {'name': 'get weather', 'args': {'city': 'Chennai'}, 'id': 'fc 8ee481c4-db27-4a4f-a4e8-839700ea29bb', 'type': 'tool call'} , invalid tool calls = , usage metadata={'input tokens': 161, 'output tokens': 85, 'total tokens': 246, 'output token details': {'reasoning': 57}} , {'role': 'tool', 'content': 'The weather in Chennai is sunny and 25°C.', 'tool call id': 'fc 8ee481c4-db27-4a4f-a4e8-839700ea29bb'} Wait Final....