Best Free AI Open Source Models tools

A lightweight, high-performance LLM proxy for caching, failover, cost tracking, and seamless integration between local and cloud AI providers.

Wolbarg is a local-first TypeScript SDK providing shared semantic memory for AI agents using SQLite or PostgreSQL.

Unified, zero-allocation native AI inference runtime for Java, consolidating LLM, vision, and speech pipelines via Project Panama.

Osaurus is a private, local AI assistant for Mac that runs open models on-device, with optional cloud AI access and autonomous agents.

A lean, fully-tested LLM inference server built for cheap CPU hardware, with persistent caching and an OpenAI-compatible API.

A dependency-free C engine that streams expert weights from disk to run the 744B-parameter GLM-5.2 MoE model on consumer hardware with as little as 25 GB of RAM.

Frugon is a free, open-source LLM cost analyzer that identifies where your LLM bill leaks by analyzing your call logs locally.

A branchless, zero-jitter ingress router for 32-GPU distributed mesh networks utilizing JAX/XLA and NCCL.

Skeights serializes fitted scikit-learn models to safetensors and JSON, replacing insecure pickle with a safe, inspectable format.

Run any LLM locally on your Mac faster than LM Studio or Ollama, with chat, coding agents, images, video, and voice, all fully offline and open source.

A lightweight, high-performance LLM proxy for caching, failover, cost tracking, and seamless integration between local and cloud AI providers.

Wolbarg is a local-first TypeScript SDK providing shared semantic memory for AI agents using SQLite or PostgreSQL.

Unified, zero-allocation native AI inference runtime for Java, consolidating LLM, vision, and speech pipelines via Project Panama.

Osaurus is a private, local AI assistant for Mac that runs open models on-device, with optional cloud AI access and autonomous agents.

A lean, fully-tested LLM inference server built for cheap CPU hardware, with persistent caching and an OpenAI-compatible API.

A dependency-free C engine that streams expert weights from disk to run the 744B-parameter GLM-5.2 MoE model on consumer hardware with as little as 25 GB of RAM.

Frugon is a free, open-source LLM cost analyzer that identifies where your LLM bill leaks by analyzing your call logs locally.

A branchless, zero-jitter ingress router for 32-GPU distributed mesh networks utilizing JAX/XLA and NCCL.

Skeights serializes fitted scikit-learn models to safetensors and JSON, replacing insecure pickle with a safe, inspectable format.

Run any LLM locally on your Mac faster than LM Studio or Ollama, with chat, coding agents, images, video, and voice, all fully offline and open source.

AXIOM is a bootable Rust kernel that optimizes transformer inference by replacing generic OS abstractions with inference-specific primitives.

NanoEuler is an open-source GPT-2-style language model built entirely from scratch in C/CUDA with hand-written backprop, BPE tokenizer, FlashAttention, and training pipelines.

Automated end-to-end few-shot text classification pipeline supporting 50+ languages, built on SetFit and SBERT embeddings.

Oxlo.ai is a privacy-first AI inference API offering request-based pricing for over 45 open-source models.

PII GUI is a local-first desktop application for detecting and redacting personal information in documents using regex and on-device ONNX models.

A Go-native framework for building, orchestrating, and observing AI agents with native OpenTelemetry observability and a self-hosted dashboard.

Ollama is a platform for running large language models locally and scaling to the cloud, offering access to faster, larger models with parallel requests and real-time web information.

Jan is an open-source, local-first AI chat app that works as a private ChatGPT replacement.

A browser-based tool that estimates which local AI models your machine can run based on device capabilities.

Fireworks AI is a generative AI platform for fast inference, model hosting, fine-tuning, and scalable deployment of open models.

PixaryAI is an online AI video generator for creating videos from text prompts or images with browser-based controls.

A lightweight, high-performance LLM proxy for caching, failover, cost tracking, and seamless integration between local and cloud AI providers.

Wolbarg is a local-first TypeScript SDK providing shared semantic memory for AI agents using SQLite or PostgreSQL.

Unified, zero-allocation native AI inference runtime for Java, consolidating LLM, vision, and speech pipelines via Project Panama.