What Neuton does

Neuton is a small, privately held AI/ML company (founded 2021, headquartered in Pleasanton, California, 11-50 employees) that has developed a proprietary neural network framework purpose-built for ultra-low-power, on-device edge AI inference on CPU-based microcontrollers and SoCs. The framework centers on a differentiated "tiny self-growing" architecture that automatically grows and optimizes during training to minimize memory footprint, paired with automated feature extraction, feature selection, and signal windowing preprocessing tools. It targets time-series and sensor data workloads, with specific support for use cases including gesture recognition, wake word detection, anomaly detection, human activity recognition, asset tracking, health monitoring, and smart home interaction.

The company's commercial footprint is almost entirely defined by a deep strategic partnership with Nordic Semiconductor: Neuton's framework is embedded as the exclusive neural network offering within Nordic's Edge AI Lab platform, supporting the full Nordic SoC lineup (nRF91, nRF54, nRF53, nRF52 series) and extending to Axon NPU-equipped devices via a no-code LiteRT Model Builder. The platform is delivered as a self-serve web product (app.lab.nordicsemi.com), with no public API, and pricing is not disclosed — commercial terms appear to be handled through direct engagement, suggesting a licensing or usage-based arrangement with Nordic rather than a broad standalone SaaS motion. Neuton's go-to-market is product-led growth routed through Nordic's developer ecosystem (Edge AI Lab, DevZone, documentation, webinars), with no independent sales infrastructure indicated in the available data.

Name

Neuton

Legal name

Neuton

Website

https://neuton.ai

Company type

Private

Founded year

2021

Operating status

Operating

Headcount range

11–50 employees

Short description

Neuton is a private AI company that has developed a proprietary neural network framework for ultra-low-power, on-device edge AI inference on CPU-based microcontrollers. It serves IoT device manufacturers and embedded developers via a deep integration with Nordic Semiconductor's Edge AI Lab platform.

Ownership category

akta.pro rank

Product category

Edge AI Neural Network Framework

NAICS

Computer Systems Design Services (541512)

SIC

Services-Computer Programming Services (7371)

akta.pro primary industry

On-Device Inference Runtimes & SDKs (mobile/embedded) (HDAAAJAB)

akta.pro secondary industries

Industry-Specific Low-Code Platforms (Vertical App Platforms) (HDAEAKAL), AI Accelerators (GPUs/TPUs/NPUs/ASICs) (HDAAAAAA)

Keywords

• Edge AI inference
• Neural network framework
• Tiny ML models
• On-device AI
• Microcontroller machine learning

Headquarters

HQ city

Pleasanton

HQ country

United States

HQ region

North America

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Infrastructure, Marketing or Sales

Revenue model

1. Edge AI Model Platform Services: Neuton provides its neural network framework through Nordic Semiconductor's Edge AI Lab platform, likely generating revenue through platform licensing or usage-based models for AI model building and deployment services.

Go-to-market motion1 record

Distribution channels2 records

Marketing channels3 records

Core offering

Neuton develops a neural network framework that automatically builds and trains tiny machine learning models for deployment on microcontroller-based edge devices. The framework is distributed through Nordic Semiconductor's Edge AI Lab platform, enabling no-code creation of ultra-low-power, on-device AI models for time-series, sensor, and audio data applications on Nordic SoC hardware.

Product overview

Neuton is a neural network framework for edge AI, available as part of the Nordic Edge AI Lab platform. The offering consists of two main components: the Neuton neural network framework for building tiny self-growing models that run on CPU of any Nordic System-on-Chip (SoC), optimized for time-series and sensor data; and LiteRT model support for running models on Nordic SoCs featuring an Axon Neural Processing Unit (NPU), enabling audio and advanced AI applications. The platform provides an automated, no-code workflow for model creation, training, and on-device deployment across Nordic hardware (nRF91, nRF54, nRF53, nRF52 series).

Differentiator

Problem solved

Functional benefit

Brands

• Neuton models: Custom neural network models for CPU-run edge AI on Nordic SoC devices
• Neuton neural network framework

Products and services

• Neuton Neural Network Framework A neural network framework for building tiny self-growing AI models that run on the CPU of any Nordic System-on-Chip (SoC). It is optimized for time-series and sensor data processing with ultra-low-power consumption for always-on sensing applications, targeted at IoT device manufacturers and edge AI developers.
• Nordic Edge AI Lab A web-based platform for building and deploying edge AI models on Nordic Semiconductor devices. Provides automated model creation tools, data preprocessing (windowing, feature extraction, selection), model analysis dashboards, and deployment capabilities across nRF91, nRF54, nRF53, and nRF52 SoC series, for IoT and embedded AI developers.
• LiteRT Model Builder A no-code model builder for configuring LiteRT models to run on Nordic SoCs featuring an Axon Neural Processing Unit (NPU). Enables audio and advanced AI applications with optimized pipelines for NPU target platforms, for developers building on Axon NPU hardware.
• Wake Word and Keyword Spotting Models Production-ready wake word and keyword spotting models built without requiring user data, created through a four-step process of entering the wake word, training the model, testing in browser, and deploying to device. Designed for voice activation on low-power Nordic devices.

Named customers1 record

Segments1 record

Ideal customer profiles1 record

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

AI capability7 records

Feature6 records

Partnerships

One partnership is on record.

• Nordic SemiconductorcoreStrategic or Co-development PartnerNordic Semiconductor integrated Neuton's neural network framework as the foundation of its Edge AI Lab platform. Neuton models are the primary offering for CPU-run edge AI on Nordic SoC devices, enabling developers to build and deploy tiny self-growing AI models. The partnership positions Neuton as the exclusive neural network framework within Nordic's edge AI ecosystem spanning multiple device generations.

Scale indicators1 record

Recent moves5 records

Expansion highlights4 records

Direct peers

• Edge Impulse: Edge Impulse offers a developer platform for building, training, and deploying ML models on edge devices (MCUs and CPUs). It is the closest direct competitor to Neuton, targeting the same IoT/embedded developers with a no-code, sensor-data-focused workflow and competing for the same on-device inference use cases.
• SensiML (QuickLogic): SensiML provides an end-to-end AutoML platform for edge AI on microcontrollers and sensors, with a focus on tiny models and time-series data. It is a direct functional competitor to Neuton in the IoT edge AI toolchain space.

Broad incumbents

• STM32Cube.AI (STMicroelectronics): STMicroelectronics bundles its own on-device AI toolchain (STM32Cube.AI) with its STM32 MCU portfolio. As a chipmaker-owned framework, it competes with Neuton for embedded AI developers and represents the bundled-software threat to independent frameworks.
• TensorFlow Lite Micro (Google): TensorFlow Lite Micro is Google's open-source framework for running ML models on microcontrollers. It is a broad incumbent competing for the same embedded ML use cases as Neuton, with the backing of a major AI ecosystem and broad hardware support.
• Qualcomm AI Hub: Qualcomm provides edge AI model deployment tooling tightly integrated with its Snapdragon and IoT chipsets. It is a broad incumbent in on-device AI that competes for the same edge inference workloads Neuton addresses.

Emerging players

• Syntiant: Syntiant builds ultra-low-power neural decision processors and offers accompanying ML tooling for always-on audio and sensor applications. It overlaps with Neuton in the wake-word and edge audio use case but is more hardware-centric.
• Deeplite: Deeplite provides automated model optimisation (compression, quantisation) for edge and embedded deployments. It is an emerging player in the on-device AI toolchain that competes for the same developer mindshare as Neuton.
• Latent AI: Latent AI focuses on efficient on-device ML inference for edge hardware, with tooling around model compression and adaptive inference. It competes with Neuton in the developer-facing edge AI deployment space.
• Plumerai: Plumerai provides on-device AI for cameras and sensors, including people detection and tracking on low-power devices. It is an emerging competitor intersecting with Neuton's edge AI use cases, particularly in vision and IoT sensing.
• BrainChip (Akida): BrainChip offers a neuromorphic processor IP and toolchain for ultra-low-power edge AI. It is an emerging player in the on-device AI hardware/software stack that competes with Neuton on always-on, low-power inference workloads.

Market position

Strengths4 records

Weaknesses4 records

Competitive moat4 records

Key risks5 records

Key highlights6 records

Customer concentration

• Xx.com/neutonai

Revenue estimate

Valuation estimate

Number of profiles

Profiles1 record

Funding overview

Funding rounds

Investors

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Neuton