cd /news/large-language-models/day-4-llm-tool-calling · home › topics › large-language-models › article
[ARTICLE · art-140233] src=dev.to ↗ pub= topic=large-language-models verified=true sentiment=· neutral

Day 4 - LLM - Tool Calling

A developer published a walkthrough of LLM tool calling, explaining that tools are functions a model can invoke to reach external APIs, databases, or knowledge bases, and that the model itself only selects which tool to call for a given query. The post demonstrates the approach in LangChain using the @tool decorator and a tool_map dictionary, with Groq as the inference provider for models such as llama3 and gpt-oss-120b, and notes MCP as the next level. It lists real-time data access, reduced hallucinations, and extended LLM capability as benefits, against token cost, latency, and security risks as drawbacks.

by read1 min views1 publishedSep 26, 2026

#

Tool Calling

  • Asking some query to the LLM.

  • LLM is like a phone book , it can't call on its own. With some tools only it will.

  • Tools are nothing but FUNCTIONS.

  • Tool calling is also called as FUNCTION Calling.

  • Its a method , which models reliably connect and interact with external tools like API , database or knowledge base.

  • LLM gets set of tools and it decides which tools needs to be invoked for a specific user query and to complete a given task.

  • Next level of this is " MCP ".

#

Code

  • I am just Mocking the data.

  • In Langchain, tool is there . Tool is like the method.its like a rapper functionality.

  • If we write any method and mention "@tool " means --> its calledDECORATOR . Its like tool and object.

  • tool_map is a dictionary.

  • ChatGroq --> if langchain is integrated with Groq.

#

Notes

  • Functions are nothing but TOOLS.
  • llm will have many functions.
  • Infra provider for LLM is Groq site . Creat API key. This is aGroq client --> it will communicate to cloud LLMs. Eg., llama3 , gpt-oss120 billion model, for this we needs API key.

#

Implementation

#

Advantages

  • Real-time data access and information retrieval
  • Reducing Hallucinations
  • Extends capability of LLMs

#

Disadvantages

  • Token consumption and Cost
  • Cost & Latency
  • Security & Safety Risks

#

Questions

source & further reading
dev.to — original article I Built a Snowflake RAG Assistant That Actually Works in Production I built a persistent context system for Claude Code using Obsidian. I built a persistent context system for Claude Code using Obsidian.
── more in #large-language-models 4 stories · sorted by recency
── more on @langchain 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/day-4-llm-tool-calli…] indexed:0 read:1min 2026-09-26 · —