ProjectNew

An operating layer for research & development.

ProjectNew brings AI reasoning, project knowledge, experimentation, code execution, and structured collaboration into a single local-first workspace.

Workspace
Local-first

Projects

Research plan

References

Knowledge base

Papers

Draft

AI

Assistant

Why ProjectNew

Built around research that stays under your control.

01

Local

All local model deployment for sensitive research work.

02

Private

Secure, internet-less environment. Your data stays with you on your own infrastructure.

03

Domain Agnostic

For all scientific streams. Support scientists, engineers, academics, and product teams in the same environment.

Key Features

01

Zero Cloud, Zero IP Leakage

Runs completely offline on local hardware with open-weight models (Granite, Gemma, Qwen, etc.).

02

Agentic Orchestration

Built-in multi-agent workflows (LangGraph + FastAPI) managing literature retrieval, statistical analysis, and experiment logic.

03

Full R&D Lifecycle

Embedded code execution (Monaco/Python), data analysis (NumPy/SciPy), vector search (ChromaDB), and automated LaTeX scientific paper compilation.

The research workspace

One system. Specialized workspaces.

ProjectNew gives research teams a single environment for planning, reading, writing, experimenting, and collaborating without losing context between tasks.

Workflow
Project
References
Papers
Chat
Iterations

Home

Primary landing interface, quick metrics, and workspace navigation overview.

Project Dashboard

Centralized overview for tracking, organizing, and managing research projects.

Reference Dashboard

Knowledge base for storing, annotating, and citing literature & references.

Paper Dashboard

Full-featured LaTeX editor and manuscript authoring workspace.

Iteration Dashboard

Phase and milestone execution workspace for iterative R&D cycles.

Chat Panel

Agentic assistant interface for code operations, tool execution, and document querying.

Specifications

Hardware Requirements & Model Architecture

Minimum Hardware Requirements

CPU Only:Minimum 16GB RAM
Small GPU:Minimum 8GB RAM + 4GB VRAM

Supported Model Family Architectures

IBM Granite Series:Granite 4.1
Google Gemma Series:Gemma 4
Qwen Series:Qwen 3.5, Qwen 3.6
Format:Only GGUF Format supported

Download

Get started with ProjectNew

Release Notice

Version 1.1.0 coming soon

The preview version 1.0.0 has been taken down. Version 1.1.0 will be made available soon with improved local inference and stability.

Windows

v1.1.0 Soon

Ubuntu

In development

Fedora

Planned

Linux Mint

Planned

Pop!_OS

Planned

Arch

Planned

Pricing

Flexible deployment from personal laptops to air-gapped research clusters.

ProjectNew is available across four clear deployment editions so individual researchers, labs, universities, and enterprise teams can adopt the same platform at the right scale.

Edition

Individual Edition

Free Forever

For solo researchers and students on PCs or laptops.

Edition

Lab Suite: Trial

Free Evaluation

For labs validating multi-agent workflows and shared knowledge bases.

Edition

Lab Suite: Academic

Subsidized / Grant-Friendly

For universities and public research institutes adopting lab-scale deployments.

Edition

Lab Suite: Enterprise & Govt.

Custom Commercial

For defense, pharma, semis, energy, and highly regulated R&D environments.

Feedback

Help shape ProjectNew.

Share your experience, ideas, and feedback through the ProjectNew form.