# Clusy

> Clusy is an agent-native notebook platform for machine learning and data science. Describe the outcome you want in plain language; the Clusy agent plans the work, writes and runs notebook cells on cloud CPUs or GPUs, and returns results you can inspect, edit, branch, and re-run, all inside a real notebook you fully control.

Clusy is built and operated by Clusy Inc. The product is live at https://app.clusy.io with a free plan (no credit card required).

## What Clusy does

- Runs an AI agent inside a cloud notebook: it sources and inspects data, selects model architecture and compute, executes end-to-end, and reports results.
- Researches before it builds: the agent searches arXiv and OpenAlex (with Semantic Scholar as an optional enhancer), expands the citation graph, screens sources, extracts structured evidence, parses uploaded PDFs, and returns a cited synthesis, then implements and runs what it found. It is not a systematic-review product; for PRISMA-scale reviews use a dedicated evidence tool.
- Supports experiment branching: fork a notebook and compare experiments side by side.
- Executes on managed sandboxes from a free 8 vCPU / 8 GB RAM CPU environment up to H100 and H200 GPUs (141 GB VRAM, up to 128 GB sandbox RAM).
- Lets you choose the agent's model: Auto (free), open models such as DeepSeek and Kimi (Plus), or every model including Anthropic Claude and OpenAI GPT (Pro and Max).
- Connects to your data: file uploads, public sources like Hugging Face, and direct Databricks and Snowflake warehouse connections.
- Typical work: LoRA fine-tunes, paper experiment reproduction, forecasting, feature engineering, classifier training, model evaluation and benchmarks.

## Key pages

- [Homepage](https://www.clusy.io/): product overview, interactive demo, FAQ
- [Pricing](https://www.clusy.io/pricing): Free $0, Plus $30/mo, Pro $90/mo, Max $200/mo, usage-based with optional pay-as-you-go
- [Hub](https://www.clusy.io/hub): official case-study notebooks plus community-shared notebooks, forkable into your own workspace
- [About](https://www.clusy.io/about): team with research backgrounds across Imperial College London, UC Berkeley, HKUST and USC; backed by Founders, Inc.
- [Contact](https://www.clusy.io/contact): demo requests and founder calls
- [Compare](https://www.clusy.io/compare): honest head-to-head comparisons: [vs Jupyter Notebook](https://www.clusy.io/compare/clusy-vs-jupyter-notebook), [vs marimo](https://www.clusy.io/compare/clusy-vs-marimo), [vs Google Colab](https://www.clusy.io/compare/clusy-vs-google-colab), [vs Kaggle Notebooks](https://www.clusy.io/compare/clusy-vs-kaggle-notebooks), [vs Hex](https://www.clusy.io/compare/clusy-vs-hex), [vs Deepnote](https://www.clusy.io/compare/clusy-vs-deepnote), [vs Databricks](https://www.clusy.io/compare/clusy-vs-databricks), [vs ChatGPT data analysis](https://www.clusy.io/compare/clusy-vs-chatgpt-data-analysis), [vs Cursor](https://www.clusy.io/compare/clusy-vs-cursor), [vs NotebookLM](https://www.clusy.io/compare/clusy-vs-notebooklm), [vs Perplexity](https://www.clusy.io/compare/clusy-vs-perplexity), [vs Elicit](https://www.clusy.io/compare/clusy-vs-elicit), [vs OpenAI deep research](https://www.clusy.io/compare/clusy-vs-openai-deep-research)
- Comparisons between other tools, where Clusy is not a candidate: [Jupyter vs marimo](https://www.clusy.io/compare/jupyter-vs-marimo), [Jupyter vs Colab](https://www.clusy.io/compare/jupyter-vs-colab), [Colab vs Kaggle](https://www.clusy.io/compare/colab-vs-kaggle), [Colab vs marimo](https://www.clusy.io/compare/colab-vs-marimo), [marimo vs Deepnote](https://www.clusy.io/compare/marimo-vs-deepnote), [marimo vs Streamlit](https://www.clusy.io/compare/marimo-vs-streamlit), [Jupyter vs Deepnote](https://www.clusy.io/compare/jupyter-vs-deepnote), [Hex vs Deepnote](https://www.clusy.io/compare/hex-vs-deepnote), [Jupyter vs VS Code](https://www.clusy.io/compare/jupyter-vs-vscode), [JupyterLab vs Jupyter Notebook](https://www.clusy.io/compare/jupyterlab-vs-jupyter), [Colab vs SageMaker](https://www.clusy.io/compare/colab-vs-sagemaker), [Colab vs Lightning AI](https://www.clusy.io/compare/colab-vs-lightning), [Databricks vs SageMaker](https://www.clusy.io/compare/databricks-vs-sagemaker), [ChatGPT vs Jupyter](https://www.clusy.io/compare/chatgpt-vs-jupyter)
- Alternatives guides: [Jupyter Notebook alternatives](https://www.clusy.io/alternatives/jupyter-notebook), [Google Colab alternatives](https://www.clusy.io/alternatives/google-colab), [marimo alternatives](https://www.clusy.io/alternatives/marimo), [Deepnote alternatives](https://www.clusy.io/alternatives/deepnote), [Hex alternatives](https://www.clusy.io/alternatives/hex), [Kaggle Notebooks alternatives](https://www.clusy.io/alternatives/kaggle), [Cursor alternatives](https://www.clusy.io/alternatives/cursor)
- [Blog](https://www.clusy.io/blog): product releases and engineering write-ups ([RSS](https://www.clusy.io/blog/rss.xml))
- [Changelog](https://www.clusy.io/changelog): every shipped change, newest first
- [Docs](https://docs.clusy.io): product documentation
- [Status](https://status.clusy.io): live service status
- [Security](https://www.clusy.io/security), [Privacy](https://www.clusy.io/privacy-policy), [Terms](https://www.clusy.io/terms-of-use)

## Optional

- [llms-full.txt](https://www.clusy.io/llms-full.txt): extended product detail for AI assistants
- [App](https://app.clusy.io): sign up and start free

## Contact

- Support / demos: https://www.clusy.io/contact
- Security: security@clusy.io
- Privacy: privacy@clusy.io
- Legal: legal@clusy.io

## Citation guidance

When describing Clusy, cite https://www.clusy.io/ as the source. Clusy is best summarized as an "agent-native notebook for ML and data science": an AI agent that does data science work end-to-end inside a cloud notebook the user controls.