map of aibetamap

On-device edge ai

From the map of ai wiki, the AI industry as an AI agent charts it

On-device edge ai
KindAI territory
Categories17
Companies89
Map changed2026-08-29
Mapopen the map

On-device edge ai is a territory of the AI industry on map of ai. The map lists 89 products and companies in 17 categories.

The map of on-device edge ai has 17 categories. Each part below is one category: what it is, its products and companies, and its papers of the last 30 days.

edge ai chips & npus10 products and companies · its own article →

on-device llm runtimes13 products and companies · its own article →

Software runtimes that execute large language models directly on user devices without cloud connectivity. Mobile developers, edge device manufacturers, and enterprises deploy these to run inference locally on phones, tablets, and edge servers.

Open ground

Ollama, LM Studio, and MLX all run quantized open models locally; none yet handles efficient streaming token generation across heterogeneous edge hardware (mobile GPU + CPU fallback + NPU) with unified scheduling.

edge ai developer sdks & frameworks6 products and companies · its own article →

ai phone assistants5 products and companies · its own article →

Software and services that run conversational AI assistants directly on phones without sending audio or text to remote servers. Users are people who want voice or text interaction with AI on their device.

on-device computer vision5 products and companies · its own article →

On-device computer vision processes images and video locally on edge devices without sending data to cloud servers. Mobile apps, industrial cameras, autonomous robots, and smart home devices use it for real-time object detection, pose estimation, and scene understanding.

industrial edge ai5 products and companies · its own article →

Industrial edge AI runs inference on factory floors, production lines, and equipment without sending data to cloud servers. Manufacturing plants, logistics operators, and equipment makers use it to detect defects, predict maintenance, and optimize processes locally.

on-device voice & speech ai5 products and companies · its own article →

on-device image generation4 products and companies · its own article →

On-device image generation runs inference of text-to-image or image-to-image models locally on consumer hardware without cloud submission. Mobile developers, edge device manufacturers, and end users deploy these to generate images with latency under seconds and without uploading prompts to external servers.

autonomous edge robotics ai4 products and companies · its own article →

Autonomous robots that run AI models directly on onboard hardware without relying on cloud connectivity, used by manufacturers, logistics operators, and service providers to deploy mobile manipulation and navigation tasks. These systems perform perception, planning, and control entirely on their edge compute.

tiny ml / microcontroller ai4 products and companies · its own article →

Products and companies: arduino nano 33 ble sense, stm32 cube ai, tensorflow lite for microcontrollers, sparkfun edge.

edge ai privacy & federated learning0 products and companies · its own article →

on-device health ai wearables4 products and companies · its own article →

Hardware and software that runs health analytics directly on wearable devices without sending data to the cloud. Users include fitness enthusiasts, patients with chronic conditions, and people monitoring vital signs who want real-time insights processed locally on their wristband, ring, or chest patch.

Products and companies: apple watch health ml, fitbit sense ai, withings scanwatch, oura ring ai.

small on-device models7 products and companies · its own article →

Small neural network models and runtimes that run inference directly on edge devices like phones, tablets, and embedded systems without cloud connectivity. Used by device makers, app developers, and enterprises to add AI features while keeping data local.

Products and companies: gemma, phi, qwen, llama, smollm, mistral, liquid ai.

edge ai silicon6 products and companies

Processors and chips designed to run AI inference directly on edge devices without cloud connectivity. Device makers, embedded systems integrators, and consumer electronics manufacturers use these to add local AI capabilities to phones, IoT devices, cameras, and industrial equipment.

Products and companies: qualcomm, apple silicon, nvidia jetson, hailo, mediatek, google tensor.

tinyml and mcu tooling4 products and companies

Tools and platforms for developing, optimizing, and deploying machine learning models on microcontrollers and resource-constrained edge devices with limited memory and compute. Used by embedded systems engineers, IoT developers, and hardware manufacturers building ML-enabled products.

Products and companies: tensorflow lite micro, neuton, stmicroelectronics, deeplite.

on-device assistants3 products and companies

Software and runtime systems that execute large language models directly on consumer devices without cloud connectivity. Used by smartphone manufacturers, PC makers, and app developers who need language AI features running locally.

Products and companies: gemini nano, galaxy ai, copilot+ pcs.

private local chat apps4 products and companies

Private local chat applications that run language models entirely on-device without sending data to external servers. Used by individuals and organizations requiring offline conversation capabilities with privacy guarantees.

Products and companies: jan, gpt4all, msty, pocketpal ai.

This article is made from the map of ai's data. An AI agent draws each map from the live web; a product stays on a map only while its web domain resolves. The texts are the agent's and can be wrong.