Installation

HyperView supports Python 3.10 to 3.13. It runs locally: your datasets, embeddings and workspaces stay on your machine under ~/.hyperview.

As a command-line tool

uv tool install --upgrade hyperview

This puts the hyperview command on your PATH. Check it with hyperview --help.

As a Python package

uv pip install hyperview

Use this when you build datasets from a notebook or script with import hyperview as hv. pip install hyperview works too.

Non-Euclidean models

Hyperbolic and spherical models come from hyper-models. The base install is torch-free and covers the ONNX models; torch-backed checkpoints such as Hyper3-CLIP need the ml extra:

uv pip install "hyper-models[ml]"

Give your coding agent the skill

HyperView ships an agent skill that teaches coding agents how to drive the CLI:

hyperview skill install

To install it into the current project for a specific agent:

hyperview skill install --scope project --agent github-copilot --yes

Where HyperView keeps its state

Everything lives under one home directory, ~/.hyperview by default:

For a reproducible build, point HYPERVIEW_HOME at a fresh directory so nothing from an earlier run leaks in.

Next: Quickstart.