Aquin

Vibe Code LLMs (use ai to make ai)

Visit website

FreeAI Coding2,941 impressions#4 of its week

>log in to comment

Aquin is a local-first platform that lets individuals and teams vibe-code their own private, secure LLMs. Users can upload data, choose open-source or hosted models, fine-tune or build RAG pipelines, and use everything transparently. Aquin removes ML and infrastructure complexity so teams can quickly build, own, and deploy domain-specific AI systems.

AI Coding

Aquin is a local-first platform for building, fine-tuning, and deploying private, secure LLMs with transparent RAG pipelines.

for: ML researchers, data scientists, and engineering teams building domain-specific AI systems.

Key features

5 features of Aquin

Local-first environment — Runs all training and inference on your own hardware, keeping data and models private.
Model selection — Choose from open-source or hosted models for fine-tuning or RAG pipelines.
AQIT CLI & SDK — Open-source command-line and Python SDK to declare recipes, train, inspect, and patch models locally.
Interpretability tools — Provides attribution, live training watch, weight editors, and circuit extraction for debugging.
Framework integration — Built as a developer environment similar to Next.js, supporting libraries, frameworks, and autonomous agents.

Use cases

Fine-tune a domain-specific LLM on proprietary data without leaving your premises
Debug failed fine-tuning runs using live training visualizations and attribution tools
Create RAG pipelines that combine private documents with open-source models
Experiment with weight-level interventions and circuit extraction for research
Deploy secure AI services for regulated industries such as life sciences or medical imaging