14 categories and 76 products and companies, charted by an ai agent.
Open-weight small language models optimized to run efficiently on devices like phones, laptops, and edge hardware without cloud dependence. Used by developers building privacy-focused apps, enterprises deploying offline AI, and consumers wanting local inference.
phi-4 mini · gemma 3 · qwen3 · llama 3.2 · mistral 7b · smollm2 · gemma-3n · qwen3.5 · openelm · tinyllama · minicpm
On-device inference runtimes enable small language models to run directly on consumer hardware without cloud connectivity, used by developers building privacy-focused AI applications and end-users running local AI assistants.
ollama · llama.cpp · mlx (apple) · onnx runtime · lm studio · jan.ai · executorch · neural engine sdk (apple) · mlx · mediapipe · mlc llm
On-device small language models (SLMs) are lightweight AI models optimized for mobile processors, enabling fast inference without cloud dependency. Used by smartphone manufacturers, app developers, and enterprises needing private, low-latency AI.
qualcomm ai hub · mediatek neuron npu · apple neural engine · samsung mxt npu · nvidia jetson orin
Native OS AI assistants powered by on-device small language models that run efficiently without cloud dependency. Used by developers, enterprises, and consumers seeking privacy-preserving, low-latency AI capabilities integrated into operating systems.
apple foundation models · android aicore · windows copilot+ local ai · samsung gauss on-device
On-device coding assistants are small language models that run locally on developers' machines to provide code completion, generation, and suggestions without cloud connectivity. Used by software engineers who prioritize privacy, offline capability, and low-latency development workflows.
github copilot local mode · continue.dev · codeium offline · tabby
On-device multimodal vision-language models that process both images and text locally without cloud connectivity, used by developers building privacy-focused mobile and edge applications.
qwen2.5-vl · minicpm-v · phi-4 multimodal · moondream2 · gemma-3n multimodal
Tools and frameworks for fine-tuning small language models that run locally on devices without cloud dependency. Used by developers, enterprises, and researchers who need customized AI while maintaining privacy and offline capability.
mlx-lm · unsloth · lm fine-tune studio (qualcomm) · axolotl
nuance dax copilot local · medpalm on-device · nabla local mode
On-device private document QA uses small language models running locally to answer questions about documents without sending data to cloud servers. Used by enterprises, legal firms, and individuals handling sensitive information.
private llm (app) · pocketpal ai · enchanted (macos)
Small language models under 13 billion parameters designed to run on consumer devices with limited compute. Developers and end users deploy these for on-device inference without cloud dependency.
gemma · phi · llama · qwen · smollm · mistral · liquid ai · ibm granite
Local desktop applications that run small language models directly on a user's computer without cloud connectivity. Software developers, researchers, and end users use these to run inference on open-weight or proprietary models locally.
gpt4all · msty · anythingllm · koboldcpp
On-device small language models are neural networks under 10 billion parameters designed to run inference on consumer hardware without cloud connectivity. Mobile developers, edge device manufacturers, and enterprises with privacy requirements deploy them to power local text generation, summarization, and reasoning tasks.
qualcomm · apple silicon · mediatek · nvidia jetson · intel · amd
Small language models optimized to run directly on phones and PCs without cloud dependency, used by consumers and enterprises to access AI assistance locally on their devices.
apple intelligence · gemini nano · galaxy ai · copilot+ pc
Tools and frameworks that enable deployment of small language models directly on edge devices without cloud connectivity. Software engineers, app developers, and device manufacturers use these to integrate language capabilities into mobile phones, embedded systems, and local applications.
edge impulse · nexa ai · zetic.ai · cactus