Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

NutriGen: Personalized Meal Plan Generator Leveraging Large Language Models to Enhance Dietary and Nutritional Adherence

Saman Khamesian, Asiful Arefeen, Stephanie M. Carpenter, Hassan Ghasemzadeh

TL;DR

NutriGen introduces an LLM-driven framework for personalized daily meal planning that integrates user dietary profiles with USDA nutrition data to produce structured, nutritionally aligned meal plans. It employs modular prompt engineering and an optional retrieval-augmented mechanism to balance usability, adaptability, and accuracy, demonstrating favorable calorie-target adherence for certain models (e.g., Llama 3.1 8B and GPT-3.5 Turbo). The work provides a comparative evaluation across diverse models, highlighting processing time trade-offs and limitations such as token-length constraints and variability in nutrition estimates, and outlines concrete future directions including real-time data retrieval, interactive interfaces, and multimodal inputs. Collectively, the findings support the feasibility of scalable, user-friendly dietary planning via LLMs while identifying practical improvements needed for deployment in real-world settings.

Abstract

Maintaining a balanced diet is essential for overall health, yet many individuals struggle with meal planning due to nutritional complexity, time constraints, and lack of dietary knowledge. Personalized food recommendations can help address these challenges by tailoring meal plans to individual preferences, habits, and dietary restrictions. However, existing dietary recommendation systems often lack adaptability, fail to consider real-world constraints such as food ingredient availability, and require extensive user input, making them impractical for sustainable and scalable daily use. To address these limitations, we introduce NutriGen, a framework based on large language models (LLM) designed to generate personalized meal plans that align with user-defined dietary preferences and constraints. By building a personalized nutrition database and leveraging prompt engineering, our approach enables LLMs to incorporate reliable nutritional references like the USDA nutrition database while maintaining flexibility and ease-of-use. We demonstrate that LLMs have strong potential in generating accurate and user-friendly food recommendations, addressing key limitations in existing dietary recommendation systems by providing structured, practical, and scalable meal plans. Our evaluation shows that Llama 3.1 8B and GPT-3.5 Turbo achieve the lowest percentage errors of 1.55\% and 3.68\%, respectively, producing meal plans that closely align with user-defined caloric targets while minimizing deviation and improving precision. Additionally, we compared the performance of DeepSeek V3 against several established models to evaluate its potential in personalized nutrition planning.

NutriGen: Personalized Meal Plan Generator Leveraging Large Language Models to Enhance Dietary and Nutritional Adherence

TL;DR

NutriGen introduces an LLM-driven framework for personalized daily meal planning that integrates user dietary profiles with USDA nutrition data to produce structured, nutritionally aligned meal plans. It employs modular prompt engineering and an optional retrieval-augmented mechanism to balance usability, adaptability, and accuracy, demonstrating favorable calorie-target adherence for certain models (e.g., Llama 3.1 8B and GPT-3.5 Turbo). The work provides a comparative evaluation across diverse models, highlighting processing time trade-offs and limitations such as token-length constraints and variability in nutrition estimates, and outlines concrete future directions including real-time data retrieval, interactive interfaces, and multimodal inputs. Collectively, the findings support the feasibility of scalable, user-friendly dietary planning via LLMs while identifying practical improvements needed for deployment in real-world settings.

Abstract

Maintaining a balanced diet is essential for overall health, yet many individuals struggle with meal planning due to nutritional complexity, time constraints, and lack of dietary knowledge. Personalized food recommendations can help address these challenges by tailoring meal plans to individual preferences, habits, and dietary restrictions. However, existing dietary recommendation systems often lack adaptability, fail to consider real-world constraints such as food ingredient availability, and require extensive user input, making them impractical for sustainable and scalable daily use. To address these limitations, we introduce NutriGen, a framework based on large language models (LLM) designed to generate personalized meal plans that align with user-defined dietary preferences and constraints. By building a personalized nutrition database and leveraging prompt engineering, our approach enables LLMs to incorporate reliable nutritional references like the USDA nutrition database while maintaining flexibility and ease-of-use. We demonstrate that LLMs have strong potential in generating accurate and user-friendly food recommendations, addressing key limitations in existing dietary recommendation systems by providing structured, practical, and scalable meal plans. Our evaluation shows that Llama 3.1 8B and GPT-3.5 Turbo achieve the lowest percentage errors of 1.55\% and 3.68\%, respectively, producing meal plans that closely align with user-defined caloric targets while minimizing deviation and improving precision. Additionally, we compared the performance of DeepSeek V3 against several established models to evaluate its potential in personalized nutrition planning.

Paper Structure

This paper contains 16 sections, 5 equations, 3 figures, 2 tables.

Table of Contents

  1. INTRODUCTION
  2. METHODOLOGY
  3. Input Data
  4. Architecture
  5. Prompt Engineering
  6. EXPERIMENTAL SETUP
  7. Dataset
  8. Selected Models
  9. RESULTS
  10. Processing Time Comparison
  11. Reported Nutritional Values vs. USDA Dataset
  12. Adherence of Meal Plans to User-Specified Targets
  13. Evaluation of Meal Plan Quality and Completeness
  14. FUTURE WORK
  15. CONCLUSIONS
  16. ...and 1 more sections

Figures (3)

  • Figure 1: The overall architecture of the NutriGen framework for personalized meal plan recommendations. The system integrates user inputs, external nutrition databases, and LLMs to generate structured and practical meal plans.
  • Figure 2: Comparison of total processing time
  • Figure 3: Comparison of the average total calories in meal plans generated by each model against the specified target for each input. The black bars represent the target values, while colored bars indicate the total calories predicted by different models. This visualization highlights the variation in model performance and their ability to generate meal plans that align with user-defined nutritional goals.