This n8n workflow turns Telegram into an AI nutritionist, positioned as an open-source alternative to paid nutrition apps like Cal AI. Send a text description, a voice message, or a photo of your food in Telegram; Gemini recognizes the meal, estimates calories plus protein, carbs, and fat, and writes everything to Google Sheets. Type “report” to receive a daily nutrition summary with visual progress bars. Its 8 nodes combine an AI Agent with Subworkflow, Merge, and Code nodes — an excellent study in the “multimodal input + spreadsheet database + custom reports” pattern.

What it does

Three logging modes: text descriptions, voice messages (auto-transcribed), and food photos (AI image analysis)
Automatic macro estimation: Gemini analyzes the food and estimates calories, protein, carbs, and fat
Goal management: set and update daily calorie/protein targets conversationally (“update my protein goal to 120g”)
Daily reports: type “report” for a same-day summary with progress bars
Two-table Google Sheets storage: a Profile table for user targets and a Meals table for per-meal logs, exportable anytime
Lightweight onboarding: simple registration that collects no personal health data (no weight or height)

Prerequisites

An n8n instance (cloud or self-hosted)
A Telegram Bot token (created via @BotFather)
Google Sheets API credentials
A Google Gemini API key (or a compatible LLM provider)

Setup

1. Import the template in n8n: https://n8n.io/workflows/7756
2. Create two Google Sheets tables:
   - Profile: User_ID, Name, Calories_target, Protein_target
   - Meals: User_ID, Date, Meal_description, Calories, Proteins, Carbs, Fats
3. Configure credentials: Telegram Bot API, Google Sheets, Google Gemini
4. Activate the workflow and send your bot a first message to register

Steps

Send the bot a food photo, voice note, or text — the AI logs calories and macros automatically
Send “report” to see today’s intake against your targets, with progress bars
Adjust goals in natural language anytime, e.g. “set my calorie target to 1800”

Use cases

Personal fitness nutrition: track meals through chat during a cut or bulk without installing another app
Replacing paid nutrition apps: self-hosted, data-owned, covering Cal AI’s core functionality
Lightweight nutritionist service: hand clients a bot link and meal logging starts immediately
Advanced n8n learning: a real-world combination of Subworkflow, Merge, Code nodes and a multimodal AI agent

Notes

Telegram and Gemini are external services; your n8n instance’s network must reach both (Gemini can be swapped for another compatible LLM)
AI calorie and macro estimates are approximations — not a substitute for professional advice in contest prep or medical diets
Voice transcription and image analysis consume more tokens than text; estimate Gemini API costs before heavy use
Diet data is personal and sensitive: prefer self-hosted n8n and restrict Google Sheets sharing

Nutrition Tracker & Meal Logger with Telegram, Gemini AI and Google Sheets (n8n)