On-device local ai
From the map of ai wiki, the AI industry as an AI agent charts it
| Kind | AI territory |
|---|---|
| Categories | 20 |
| Companies | 105 |
| Part of | neural networks |
| Papers, 30 days | 4 |
| Map changed | 2026-08-29 |
| Map | open the map |
On-device local ai is a territory of the AI industry on map of ai. In short: on-device local ai runs model inference directly on phones, laptops, wearables, and dedicated hardware instead of sending every request to remote servers. The map lists 105 products and companies in 20 categories.
papers4 in the last 30 days, the first 4
- Local AI Agents in Power Electronics Simulation: Deployment, Measurements and Costs 2026-11-18
- Superfluid: Local AI Server for Distributed LLM Inference 2026-11-13
- Hibana: Fast Agentic Inference on Unified-Memory Devices 2026-11-07
- Learning Structured Sparsity for Large Models: Bridging Model Compressibility and Efficient Inference 2026-10-22
The map of on-device local ai has 20 categories. Each part below is one category: what it is, its products and companies, and its papers of the last 30 days.
local llm inference runtimes11 products and companies · its own article →
Software runtimes that enable large language models and AI inference to run locally on devices without cloud dependencies. Used by developers, enterprises, and end-users prioritizing privacy, latency, and offline capability.
Open ground
Better standardization and interoperability across different model formats and hardware accelerators remains fragmented and difficult for non-experts.
ai pc hardware & npus6 products and companies · its own article →
Products and companies
on-device small language models7 products and companies · its own article →
Small language models optimized to run locally on devices without cloud connectivity, enabling private and fast AI inference. Used by developers, enterprises, and consumers prioritizing privacy and offline functionality.
on-device model optimization & compression6 products and companies · its own article →
local ai desktop assistants7 products and companies · its own article →
Local AI desktop assistants are AI models that run entirely on a user's computer without cloud connectivity, prioritizing privacy and offline functionality. Used by developers, privacy-conscious professionals, and organizations with data security requirements.
Products and companies
on-device coding assistants5 products and companies · its own article →
On-device coding assistants are AI models that run locally on a developer's machine without cloud connectivity, providing code completion, generation, and analysis. Used by developers who prioritize privacy, offline work, and reduced latency.
Products and companies
on-device voice & speech5 products and companies · its own article →
on-device image & vision ai6 products and companies · its own article →
Products and companies
mobile on-device ai (android & ios sdks)5 products and companies · its own article →
Products and companies
industrial & edge ai inference5 products and companies · its own article →
Products and companies
local ai knowledge & rag tools5 products and companies · its own article →
Products and companies: anything llm, privateai, open webui, quivr, khoj.
autonomous local ai agents0 products and companies · its own article →
Autonomous local AI agents are self-directed software systems that run entirely on user devices without cloud connectivity, making real-time decisions and automating tasks locally. Used by privacy-conscious professionals, enterprises with offline requirements, and edge computing operators.
on-device model families11 products and companies
Model families designed to run inference directly on user devices without cloud connectivity, used by developers and end-users to deploy language models, vision models, and other AI workloads locally. These families provide pre-trained or fine-tunable models optimized for on-device execution.
Products and companies: phi-4, gemma 3, llama 3, mistral, qwen, smollm, liquid lfm, deepseek-r1, gemma, llama, phi.
mobile os ai platforms4 products and companies
Mobile OS AI platforms are operating system-level machine learning frameworks that run inference and model operations directly on smartphones and tablets. Device manufacturers, app developers, and system integrators use them to execute AI tasks without cloud dependency.
Products and companies: gemini nano, samsung gauss, qualcomm ai hub, mediatek neuropi.
pc local ai assistants3 products and companies
Software that runs large language models and other AI models directly on a user's computer without cloud connection, used by developers, researchers, and end-users who want offline inference. These tools execute model inference on local hardware.
Products and companies: microsoft copilot+ pc, microsoft foundry local, lenovo qira.
local ai hardware3 products and companies
Hardware devices designed to run AI models locally without cloud connectivity, used by individuals and organizations that need on-device inference. These devices execute language models, vision models, and other AI workloads directly on the hardware.
Products and companies: snapdragon x elite, intel core ultra, framework laptop.
Papers, last 30 days:model optimization and quantization6 products and companies · its own article →
Model optimization and quantization reduces the size and computational requirements of neural networks through techniques like pruning, distillation, and bit-width reduction, enabling them to run on resource-constrained devices. Machine learning engineers, embedded systems developers, and mobile app makers use these tools to deploy models locally.
Products and companies: hugging face transformers, gguf / ggml, executorch, tensorflow lite, onnx runtime, ctransformers.
os-level device ai2 products and companies
Operating system-level AI that runs directly on device hardware without cloud dependency, used by OS makers and system integrators to embed intelligence into core device functions. Executes machine learning models for features like on-device search, voice processing, image recognition, and device optimization at the firmware or kernel level.
Products and companies: copilot+ pcs, galaxy ai.
ai silicon and npus5 products and companies
Silicon chips and neural processing units designed to run AI inference locally on devices without cloud connectivity. Manufacturers, device makers, and enterprises deploy these to enable on-device machine learning.
Products and companies: qualcomm, nvidia, amd, intel, mediatek.
private local transcription3 products and companies
Software that transcribes speech to text running entirely on a user's device without sending audio to external servers. Developers, accessibility users, and privacy-conscious individuals use it to convert spoken words into written text locally.
Products and companies: macwhisper, superwhisper, vibe.
See also
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. Demo page.