All Open Source Apps

A Microsoft-built training framework that optimizes AI agents with reinforcement learning, automatic prompt optimization, or supervised fine-tuning — with near-zero code changes to your existing agent, in any framework.

Give your LLM agents on-the-job training—ART lets you apply GRPO reinforcement learning to any multi-step agentic workflow with minimal code changes.

Compress everything your AI agent reads — tool outputs, logs, RAG chunks, and files — before it reaches the LLM, achieving 60–95% fewer tokens with the same answers.

An open-source context database that gives AI agents a unified filesystem for memory, resources, and skills with hierarchical tiered retrieval.

Local-first AI memory with verbatim storage, pluggable backends, and 96.6% retrieval recall on LongMemEval — no API key required.

Mesh LLM pools GPUs and memory across every machine you own into one OpenAI-compatible API, so agents tap distributed compute instead of a single GPU box or a metered cloud bill.

Drop your files and search them instantly — no vector DB, no indexing pipeline, just raw data queried by a self-evolving intelligence layer.

Visual ETL that compiles to Polars — build pipelines on a canvas, export as standalone Python, and run anywhere without platform lock-in.

Transform messy, unstructured documents into persistent, navigable memory that AI agents can actually use.

The dbt-native data observability CLI that turns your existing dbt tests and metadata into anomaly detection, lineage graphs, and Slack/Teams alerts — no separate platform required.

A self-hosted, OpenAI-compatible LLM gateway that puts one endpoint in front of 40+ providers, with virtual API keys, per-user budgets enforced before spend, and full usage tracking.

An all-in-one open-source training suite for finetuning diffusion models—FLUX, SDXL, Wan video, and audio—on consumer GPU hardware via CLI or web UI.

A universal swarm intelligence engine that spawns thousands of autonomous AI agents to simulate and predict how real-world events unfold across social, financial, and political domains.

An open source inference engine that profiles your hardware, recommends the best local models, and tunes them for the coding agent you already use.

A pretrained decoder-only foundation model by Google Research that delivers zero-shot time series forecasting with calibrated quantile prediction intervals — no training required.

SOTA open-source dual-autoregressive text-to-speech model with rapid voice cloning, inline emotion tags, and real-time streaming inference across 80+ languages.

Self-hosted, open-source email API that automates DNS, manages SMTP delivery, and provides deep observability — replacing SES, Mailgun, and SendGrid with infrastructure you fully own.

Turn AI-written code into real apps — an open-source, self-hostable backend with Database, Auth, Storage, Memory, Functions, Assets, AI Gateway, and cron built for AI coding agents.

A next-generation documentation site generator for dbt Core projects — lineage explorer, health scoring, and full-text search for teams without access to dbt Cloud's built-in docs features.

A Rust-based real-time data processing engine for AI/ML feature computation, built on Apache DataFusion and Arrow — positioned as an alternative to Flink, Spark, Chronon, and OpenMLDB with unified streaming, batch, and request-time execution.

Collaborate on high-quality AI training data with a self-hosted annotation platform built for LLMs, NLP, and multimodal models.

An Apple-Silicon-native LoRA fine-tuning tool for Gemma on text, image, and audio data — with a wizard CLI, live browser-based training visualizer, and streaming from GCS/BigQuery for datasets too large for local disk.

Rasa Open Source is a Python machine learning framework for building contextual, multi-turn chatbots and voice assistants that understand natural language and maintain conversation state.

Federated AI search and RAG across 100+ enterprise sources—no data extraction, no vector database required.