Local & Self-Hosted AI Tools

Mini SWE Agent is a command-line AI coding assistant for developers tackling GitHub issues and everyday coding tasks. Its small Python core also suits researchers who want to inspect and modify an agent for model evaluations.

jetson-containers is a Docker container build system for developers running local AI on NVIDIA Jetson. It assembles GPU software and its dependencies into containers for workloads that need a compatible CUDA stack.

Code Cortex MCP is a local MCP server that helps AI coding agents understand how a codebase fits together. It indexes functions, classes, call paths and service connections, so agents can answer relationship questions without repeatedly searching individual files. It builds on codebase-memory-mcp.

FlashAttention is an attention library for developers training or running transformer models on their own GPUs. It computes exact attention while using less memory than standard PyTorch attention, allowing longer sequences.

MOSS is a conversational language model from Fudan University for developers and researchers who want to run and adapt a local LLM. It includes model weights, dialogue datasets and fine-tuning code.

OpenLLMetry helps developers trace LLM applications alongside their existing database and API activity. It extends OpenTelemetry, so teams can send AI application traces to the monitoring system they already use.

PaddleSeg is an image segmentation toolkit for developers building computer vision applications on their own hardware. It covers annotation, training and deployment for tasks such as medical image analysis, industrial inspection and portrait effects.

Bergamot Translator is a local AI translation library for developers building translation into browser or native applications. People who want to translate web content can use it through the separate TranslateLocally extension for Firefox.

Ente Photos is a Google Photos alternative for people who want to back up and share their library without letting the storage provider read its contents. You can self-host its server or use Ente's hosted service.

Opik is an LLM observability and evaluation platform for teams building AI agents and retrieval-based applications. It helps developers inspect failures and compare behavior across development, testing and production.

OpenCLIP is a Python library for developers and researchers building image-text matching and zero-shot classification applications. It supports pretrained CLIP models and training models on your own hardware.

Liger Kernel is a Python library for developers training or fine-tuning LLMs on their own hardware. Fused kernels reduce memory use during model operations and post-training.

ggml is a library for developers building machine learning runtimes and local AI applications. It provides tensor computation across different kinds of hardware; GGUF is the model file format covered by the project.

PageLM is a self-hosted AI study platform inspired by NotebookLM. Students, educators and researchers can use it to turn their materials into resources for revision and exam practice.

NeuralForecast is an open-source Python library for data scientists and developers who want to train and compare neural time series forecasting models in their own Python environment. It brings recurrent networks and forecasting-specific architectures under one interface, so comparing models doesn't require working with separate research implementations. The license is Apache 2.0.

ExecuTorch is an on-device AI runtime for developers adding PyTorch models to apps and embedded devices. It handles inference on the user's hardware across Android, iOS, macOS, Linux and Windows, as well as embedded systems and microcontrollers.

tlm is a terminal assistant for people who want help writing shell commands or understanding commands they've found elsewhere. It uses local models through Ollama rather than a cloud AI service.

Aperty is a desktop AI portrait editor for photographers working on weddings, family sessions and other portrait shoots. It helps apply consistent retouching across a photo set while leaving control over the strength of each adjustment.

SkyPilot manages AI training, inference and development workloads across infrastructure you control. It's for AI teams sharing compute and infrastructure teams managing jobs across clusters and cloud providers.

IndicTrans2 is a family of translation models for developers building multilingual applications and researchers working on Indian languages. It covers all 22 scheduled languages of India, including languages with limited training data.

Pydantic DeepAgents is a terminal coding assistant and Python framework for developers who want to use or build their own AI agents. Both share a Pydantic AI foundation for carrying out tasks across files and tools.

DeepGEMM is an open-source CUDA kernel library for developers building LLM training and inference systems on their own NVIDIA GPUs. It supplies matrix operations and specialized computations for mixture-of-experts (MoE) models.

x64dbg-MCP Server connects an MCP-compatible AI assistant to x64dbg for direct inspection and control of a debugging session. It's for reverse engineers, malware analysts and security researchers investigating Windows binaries.

DeepTrack2 is an open-source Python library for researchers building image data pipelines for machine learning and digital microscopy. It brings image generation, manipulation and analysis into a modular framework that runs in your own Python environment. Its focus is scientific imaging, with applications in microscopy, optics and biophysics.

BAH Browser is an MIT-licensed desktop browser for people who want an AI agent to carry out web tasks in their own browser session. It runs on Windows and combines a tabbed Chromium browser with an assistant that reads pages, clicks links and fills forms in response to plain-language requests.

RAGatouille is a Python library for developers building retrieval-augmented generation (RAG) applications with ColBERT. It builds and searches document indexes on your own hardware or server, with compressed indexes stored on disk that your application can query directly.

Leon is a personal AI assistant for people who want to automate tasks on their own computer or server. Its skills let you add actions suited to your needs.

Accentize DeRoom is a proprietary desktop audio plugin for podcast editors, content creators and dialogue editors working with speech recorded in echoing rooms. It runs inside audio and video applications on macOS and Windows, using a neural network to separate direct sound from room reflections.

Usertour helps teams build in-app guidance for browser applications. It's an alternative to Appcues and Userflow for teams that want to host their onboarding platform themselves.

data-to-paper is an open-source Python framework for researchers who want AI agents to work through a dataset and draft a scientific paper whose numerical claims they can check. Its defining feature is traceability: readers can click a value in the manuscript and follow it back to the specific analysis code that produced it.

Ruflo is an AI agent framework for developers who want Claude Code and Codex to work as coordinated teams. It adds specialist roles and shared context for projects that involve coding, testing, architecture and documentation.

Mosec is a self-hosted model serving framework for developers turning trained machine learning models into APIs. It runs on Linux and macOS, on your own hardware or cloud servers, and is open source under the Apache 2.0 license. Its main focus is handling concurrent requests while making better use of CPU and GPU resources.

Phi-4 is Microsoft's language model for developers building applications that need complex reasoning and math problem solving. It can run on your own hardware without a cloud connection.

LMCache helps teams serving LLMs reuse context that a model has already processed. It reduces repeated work on long prompts, conversation history and retrieved documents, shortening the wait for the first token and improving throughput.

Revornix is an AI knowledge workspace for people who collect articles, research papers and recordings and want to ask questions across their library. It connects saved material in a personal knowledge graph and uses those relationships to inform its answers.

Tolaria is a desktop knowledge base for personal notes, company documentation, and assistant memory. It keeps your material in ordinary Markdown files that other editors and tools can read.

Voice Changer (w-okada), also called VC Client, converts speech as you talk using AI voice conversion models. It suits people who want live voice changes, including gaming commentators running other demanding applications.

ChatTTS Colab turns text into speech for people producing narration or scripts with different character voices. It builds on ChatTTS and provides a browser interface for comparing voices and generating long audio passages.

OwnOrbit AI, formerly LifeOS AI, is an alpha self-hosted personal AI assistant for people who want help finding forgotten commitments in their own notes. It reads a Markdown folder you control and can surface deadlines, renewals, and unfinished work. The assistant runs on your computer, with a paired mobile browser app for phone access.

Argos Translate translates text on your own machine for people who need offline translation and developers adding translation to their software. Once you have the language models, it can translate without an internet connection or sending text to an online service.

gollama is a terminal app for people managing a local LLM collection in Ollama. Its keyboard-driven interface helps you compare downloaded models and clear out ones you don't need.

Forge is a terminal coding assistant that works on files in your own development environment while sending model requests to a configured provider, including OpenAI or Anthropic. It's for developers who want an agent to make code changes and run commands within their existing projects.

VoxelMorph is a Python library for researchers and developers who need to align medical images using learned deformations. It provides registration networks and training components, including unsupervised methods, with PyTorch and TensorFlow backends. The code is open source under the Apache 2.0 license.

VideoCrafter generates video on your own hardware from written prompts or still images. It's aimed at researchers and people exploring video generation for personal use; the toolbox limits use to personal, research and non-commercial purposes.

RD Agent is a self-hosted AI research agent for data scientists, researchers and quantitative developers. It runs experiments on a Linux machine or server, while chat and embedding requests use providers through LiteLLM, including OpenAI and Azure OpenAI. Those remote configurations require API credentials and internet access; hosting the agent yourself doesn't move their inference onto your hardware.

Codanna helps developers and AI coding agents explore a repository and assess which code a change could affect. It connects code search with function relationships, so an assistant can get context around a match in one response.