{"slug": "show-hn-pluto-ai-model-run-tiny-mcus-hardware-control-of-natural-language", "title": "Show HN: Pluto AI model, run tiny MCUs hardware control of natural language", "summary": "Pluto is a C++-based embedded intelligence framework that runs natural-language hardware control directly on microcontrollers with roughly 0.5 MB of RAM, supporting Arduino Mega, ESP32, Raspberry Pi, and STM32 platforms. The system interprets natural-language instructions into structured operations for GPIO, PWM, I²C, SPI, UART/serial, servo, analog I/O, timers, sensors, and FRAM persistent memory, and is offered as a pre-configured ESP-based board rather than a conventional software download because its core architecture is proprietary.", "body_md": "**Pluto — Talk to your hardware instead of programming it.**\n\n**Pluto is a Natural Language system designed to run directly inside microcontrollers (MCUs) and control the hardware from within the chip itself — without requiring traditional programming for every new task.**\n\nPluto is designed to operate with an extremely small memory footprint and can run effectively with approximately **0.5 MB of RAM**.\n\nPluto currently supports the following MCU and embedded platforms:\n\n- **Arduino Mega**\n- **ESP32**\n- **Raspberry Pi**\n- **STM32**\n\nPluto is designed to be adaptable across different microcontroller and embedded hardware architectures, allowing the same natural-language control concept to be applied across multiple platforms.\n\nPluto is designed to interpret natural-language instructions and translate them into structured operations for embedded hardware.\n\nIt can understand and control:\n\n- **GPIO** — Digital HIGH / LOW input and output\n- **PWM** — Variable duty-cycle control for LEDs, motors, and other devices\n- **I²C** — Communication with sensors, displays, memory, and peripheral devices\n- **SPI** — High-speed communication with compatible peripherals\n- **UART / Serial** — Serial communication with external devices\n- **Servo** — Position control and movement sequences\n- **Analog I/O** — Reading analog sensors and generating hardware responses\n- **Timers / Frequency** — Hardware timing, pulse and frequency generation\n- **Sensors** — Processing sensor data and using it in conditional decisions\n- **FRAM / Persistent Memory** — Storing states, instructions, and persistent data\n\nPluto can be integrated directly into a microcontroller as an independent intelligence layer, allowing the device to interpret instructions and make hardware-level decisions locally.\n\nFor example, a drone could use Pluto to interpret high-level commands and coordinate sensors, motors, and other onboard hardware without relying on a cloud connection.\n\n**Natural Language → Intelligence → Decision → Hardware**\n\nPluto is designed to allow users to control hardware through natural-language instructions instead of writing traditional firmware code.\n\nThe user describes what they want to do, and Pluto interprets the instruction, organizes the required operations, and executes them through the connected MCU and hardware.\n\nSuppose a gyroscope sensor is connected to the **SDA/SCL** pins, and a servo motor is connected to **Pin 15**.\n\nThe user can simply tell Pluto:\n\n\"I connected a gyroscope sensor to your SDA and SCL pins, and a servo motor to Pin 15. Keep the servo balanced at 0 degrees.\"\n\nPluto interprets the instruction and performs the required hardware operations.\n\n```\nUser:\n\"I connected a gyroscope sensor to SDA/SCL and a servo motor\nto Pin 15. Keep the servo balanced at 0 degrees.\"\n\n        ↓\n\nPluto Understanding\n\n        ↓\n\nI²C → Gyroscope\nPin 15 → Servo Motor\n\n        ↓\n\nRead Gyroscope\n        ↓\nProcess Orientation\n        ↓\nCalculate Required Servo Position\n        ↓\nControl Servo\n```\n\n# PLUTO is a revolutionary architecture designed to operate directly on MCUs, enabling intelligent, real-time control of hardware at the embedded level. Unlike conventional systems that rely on pattern matching, Pluto is designed to emulate the structure of human thought—transforming natural language into structured reasoning and actionable instructions. This makes Pluto more than an embedded control architecture; it represents a direct step toward bringing AGI-inspired reasoning into the world of edge devices and autonomous machines.\n\nPluto is not distributed as a conventional software download.\n\nPluto is a **C++-based embedded intelligence framework** whose core architecture has been developed through several years of independent research and development. Because the core architecture is proprietary, we do not publicly release the complete internal implementation.\n\nIf you want to use the current Pluto model, we provide a **pre-configured ESP-based board with the Pluto framework installed**.\n\nAfter purchase, we prepare the hardware, install the Pluto framework, and ship the device directly to your provided address.\n\nThe listed price includes the costs associated with:\n\n- Hardware\n- Pluto framework installation and configuration\n- Development and preparation\n- Packaging and international shipping\n- Applicable taxes and related costs\nWe are currently an **independently funded research project without external investors**. Revenue from Pluto purchases helps fund continued research, development, hardware testing, and future versions of the architecture.\n\n**Customers who purchase Pluto will receive future model/framework updates for their purchased version at no additional software licensing cost**, subject to hardware compatibility.\n\nBy purchasing Pluto, you are not simply purchasing an ESP board. You are supporting an independent research project and receiving a pre-configured hardware implementation of the current Pluto architecture.\n\nPluto is available as a **physical hardware product**.\n\n**$450 USD**\n\n**$799 USD**\n\nIts core purpose is simple:\n\n**Instead of programming the microcontroller, communicate with it using natural language.**\n\nPluto is designed to be adaptable across a wide range of MCU architectures. The system can interpret natural-language instructions and convert them into structured hardware operations that the microcontroller can execute.\n\nFor persistent data storage, Pluto can work with an external **32 KB non-volatile FRAM** connected through I²C:\n\n**MB85RC256V I²C Non-Volatile FRAM**\n\nThis memory can be used to retain user-defined tasks, instructions, states, and other persistent information even when the MCU is powered off.\n\nPluto is designed to operate **without relying on third-party AI APIs or external cloud AI services**.\n\nThe goal is to keep the core intelligence and hardware-control process directly on the embedded system.\n\nPluto is not designed around conventional keyword or simple pattern-matching approaches.\n\nInstead, Pluto is being developed around a structured **natural-language understanding and reasoning approach**, inspired by how humans interpret instructions, understand relationships between actions, and organize tasks before execution.\n\nThis makes Pluto more than a simple command parser.\n\nWe see Pluto as an **early exploration toward AGI-like intelligence at the edge** — bringing natural-language understanding, task organization, and hardware control directly into embedded systems.\n\n**Natural Language → Understanding → Structured Instruction → Hardware Execution**\n\nPluto's long-term goal is to make embedded hardware programmable through **communication rather than traditional code**.\n\nPluto is currently under active development.\n\nWe are working toward **Pluto 1.5 Flash** and **Pluto 2.5**, with a major expansion of Pluto's capabilities.\n\nThe future versions are being designed to run directly on smartphones and use a portion of the phone's available system resources. Pluto 1.5 Flash and Pluto 2.5 are planned to use approximately **2 GB of RAM and 10 GB of storage** from the host smartphone, rather than requiring dedicated memory and storage hardware.\n\nFor example, on a smartphone with **8 GB of RAM**, Pluto may use approximately **2 GB of the available RAM** while running, with the remaining system resources available to the smartphone and other applications.\n\nThe goal is for Pluto to understand and process user instructions in a more human-like way and use that understanding to perform tasks.\n\nPluto will be designed to communicate directly with the smartphone and control a wide range of external hardware and embedded systems.\n\nPluto is not intended to operate primarily through simple pattern matching.\n\nThe long-term objective is to develop a system capable of **understanding information, reasoning about instructions, organizing tasks, and taking appropriate actions** in a way that is closer to human information processing.\n\nThis development represents our exploration toward **AGI (Artificial General Intelligence)** — bringing more general-purpose intelligence from software systems into real-world hardware interaction.\n\nPluto is an independent research and development project.\n\nIf you are interested in supporting the development of future Pluto versions, hardware integration, and embedded intelligence, you can support the project or get in touch with us.\n\n**Development support, collaboration, and early-access opportunities will be announced as the project progresses.**\n\n🌐 **Website:** [Visit us](https://dibsoftiot.com/)\n\n💼 **LinkedIn:** [Follow LinkedIn](https://www.linkedin.com/in/nazmul-agi/)\n\nIf you would like to support the ongoing development of Pluto:\n\n**Pluto — Talk to your hardware instead of programming it.**", "url": "https://wpnews.pro/news/show-hn-pluto-ai-model-run-tiny-mcus-hardware-control-of-natural-language", "canonical_source": "https://github.com/dib-artificial/Pluto", "published_at": "2026-10-08 09:00:55+00:00", "updated_at": "2026-10-08 09:19:34.086759+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-agents", "ai-tools", "developer-tools"], "entities": ["Pluto", "Arduino Mega", "ESP32", "Raspberry Pi", "STM32"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/show-hn-pluto-ai-model-run-tiny-mcus-hardware-control-of-natural-language", "markdown": "https://wpnews.pro/news/show-hn-pluto-ai-model-run-tiny-mcus-hardware-control-of-natural-language.md", "text": "https://wpnews.pro/news/show-hn-pluto-ai-model-run-tiny-mcus-hardware-control-of-natural-language.txt", "jsonld": "https://wpnews.pro/news/show-hn-pluto-ai-model-run-tiny-mcus-hardware-control-of-natural-language.jsonld"}}