An eight-year-old developer, Laksh, has developed a compact offline AI storytelling device capable of generating and narrating stories without requiring an internet connection.

The project is built around the ESP32-S3 microcontroller, a low-cost development board priced at approximately Rs 760, demonstrating how artificial intelligence applications can run on affordable hardware.

Works Without Internet or Cloud Services

Unlike many AI applications that depend on cloud computing, the device performs all processing locally.

The hardware includes:

• ESP32-S3 microcontroller
• INMP441 microphone
• I2S speaker
• MAX98357A audio amplifier

The system operates entirely offline and does not require:

• Wi-Fi
• Cloud services
• External APIs

This enables the device to function even in areas with limited or no internet connectivity.

AI Can Tell Stories Through Voice Commands

The storytelling device responds to simple spoken instructions such as:

• "Make a funny story"
• Play
• Pause
• Resume
• Next story

After receiving a command, the AI generates a new story, reads it aloud using text-to-speech technology and displays simple animations on its screen.

The demonstration showed that all AI processing takes place directly on the ESP32 chip without communicating with remote servers.

Built on an Open-Source AI Model

Laksh built the project by adapting TinyStories LLM for ESP32, an open-source language model originally created by developer slvDev.

The original model enables story generation on the ESP32 microcontroller without internet access.

Laksh expanded the project by adding several new capabilities, including:

• Voice interaction
• Playback controls
• Animated visual output

The original project's source code is publicly available on GitHub, allowing developers to study, modify and build upon the model.

Improving Voice Quality Is the Next Goal

According to Laksh, the primary limitation of the current prototype is its text-to-speech (TTS) output.

While the AI successfully generates and narrates stories, the spoken English voice remains relatively basic due to the hardware constraints of the ESP32 platform.

Improving voice quality is the next planned enhancement for the project.

Why This Matters

The project demonstrates how open-source AI models and affordable hardware can make artificial intelligence more accessible. Running AI directly on edge devices without cloud connectivity can reduce costs, improve privacy and enable educational tools in regions with limited internet access. It also highlights the growing potential of small language models designed for compact, low-power devices.

Frequently Asked Questions (FAQs)

Who built the offline AI storytelling device?

The project was developed by eight-year-old Laksh, co-founder of Projects by Laksh.

What hardware does the device use?

It is powered by an ESP32-S3 microcontroller, along with an INMP441 microphone, an I2S speaker and a MAX98357A amplifier.

Does the device require internet access?

No. The AI operates entirely offline without Wi-Fi, cloud services or APIs.

Which AI model powers the project?

The device is based on the open-source TinyStories LLM for ESP32, which Laksh modified with additional voice and animation features.

What improvements are planned?

Laksh aims to improve the device's text-to-speech (TTS) voice quality in future versions.