Multi Agent
For The Sciences

Powered by CMBagent, an award winning multi-agent orchestration system developed by researchers at the intersection of computational astrophysics and machine learning, and now the architectural backbone of Denario's research engine.

Quantum Physics Graphic
Install FAQ Articles Team
Hugging Face

THE PRIMARY TEAM

Francisco Villaescusa-Navarro, Ph.D

Francisco Villaescusa-Navarro, Ph.D

Research Scientist, Cosmology, CCA,
Flatiron Institute

Francisco (Paco) Villaescusa-Navarro is a computational
cosmologist and AI researcher whose work aims to transform
how scientific discovery is conducted.

Dr Boris Bolliet

Dr Boris Bolliet

Researcher

Boris was the first to introduce multi-agent systems in Astrophysics and now works across disciplines developing systems that can conduct research autonomously and make discovery faster.

Pablo Villanueva Domingo

Pablo Villanueva Domingo, Ph.D

Cosmologist, Deep Learning Scientist at Calia, Computer Vision Center, Universitat Autònoma de Barcelona

Pablo Domingo is a physicist and deep learning scientist. In 2021 he completed his PhD in Physics at the University of València and Instituto de Física Corpuscular, investigating the nature of dark matter through cosmology, specially via the 21 cm cosmological signal of the hydrogen.

VIEW WHOLE TEAM

BUILT BY RESEARCHERS FROM

Autonomous University of Barcelona MIT Harvard DeepMind Oxford Johns Hopkins Princeton Infosys Tel Aviv University University of Cambridge Flatiron Institute University of Virginia University of Chicago University of Texas at Austin

OUR PAPER

ARXIV

In this work, we describe in detail Denario and its modules, and illustrate its capabilities by presenting multiple AI-generated papers generated by it in many different scientific disciplines such as astrophysics, biology, biophysics, biomedical informatics, chemistry, material science, mathematical physics, medicine, neuroscience and planetary science

Read Arxiv Paper
Cancer cells
“

We designed Denario with a modular architecture so that users can choose which of its components best fit their research, whether that's coding, exploring research ideas, summarizing results or generating a plot.

DR. BORIS BOLLIET / UNIVERSITY OF CAMBRIDGE CAVENDISH LABORATORY
Planet Mars

PROCESS

Novelty Search

Novelty Search

Denario scans academic repositories to identify unexplored intersections between disparate fields. Outputs validated hypothesis based on existing literature.

Autonomous Coding

Autonomous Coding

Executes Python environments to process data, run statistical models, and generate visualizations. Handles dependency management and error correction automatically.

Paper Drafting

Paper Drafting

Compiles research findings into formatted LaTeX or Markdown manuscripts. Automatically manages citations, section structuring, and figure placement.

Peer Review

Peer Review

Simulates double-blinded peer review using multi agent critique. It analyzes the drafted paper for clarity, novelty, and methodological soundness, ensuring statistical rigor and adherence to field-specific standards.

ANTICIPATED CHALLENGES

Denario and CMBagent featured in the
LSST DESC AI Roadmap.

The Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST) will produce unprecedented volumes of heterogeneous astronomical data—images, catalogs, and alerts—that challenge traditional analysis pipelines. The LSST Dark Energy Science Collaboration (DESC) aims to derive robust constraints on dark energy and dark matter from these data, requiring methods that are statistically powerful, scalable, and operationally reliable.

Read more

AI GENERATED PAPERS

Characterizing the Variability and Correlates of U.S. ART Clinic Performance During the COVID-19
Pandemic (2020-2022)

Read Paper
Generated Paper

Geometric Structure of PINN Latent Space for Burger’s Equation: Low-Dimensional Manifolds and
Initial Condition Encoding

Read Paper
Generated Paper

Single-cell analysis reveals divergent developmental trajectories and regulatory networks in Plasmodium falciparum lab strains and field isolates

Read Paper
Generated Paper
View more

FREQUENTLY ASKED QUESTIONS

Is Denario free to use?

Yes, Denario is open-source software (MIT License). You can clone the repo and run it locally for free. We also offer a managed cloud version for teams that need high-availability GPUs without setup.

Which LLMs are supported?

Denario is model-agnostic. It works out-of-the-box with GPT-4, Claude 3.5 Sonnet, and local models via Ollama (Llama 3, Mistral). You can mix and match models for different agents.

Can it really execute code securely?

Yes. All code execution happens within isolated Docker containers. The system has no network access unless explicitly whitelisted, ensuring your local environment remains safe.

How does the "Peer Review" agent work?

The Peer Review agent is prompted with top-tier conference guidelines (NeurIPS, ICML). It analyzes the drafted paper for clarity, novelty, and methodological soundness, providing a detailed markdown report.

IN THE PRESS

News Nov 04, 2025

Meet Denario — An AI Assistant for Every Step of the Scientific Process

Scientists present a new type of "scientific assistant" that can synthesize papers, formulate questions, and write manuscripts.

Read Article →
Press Dec 10, 2024

The AI ‘research assistant’ that is already getting its own papers published

An international team has released an AI system capable of autonomously conducting scientific research for about $4 per paper.

Read Article →
Community Nov 20, 2025

ICCUB researchers contribute to development of Denario

Researchers from ICCUB played pivotal roles in both the technical development and ethical framing of the new AI assistant.

Read Article →
VIEW ALL PRESS

It starts with a question. It thrives in the freedom to test. It succeeds when we try to solve it together.