PyGraphistry: Leverage the power of graphs & GPUs to visualize, analyze, and scale your data

PyPI - Downloads

<table style="width:100%;"> <tr valign="top"> <td align="center"><a href="https://hub.graphistry.com/graph/graph.html?dataset=Facebook&splashAfter=true" target="_blank"><img src="https://i.imgur.com/z8SIh2E.png" title="Click to open."></a> <a href="https://hub.graphistry.com/graph/graph.html?dataset=Facebook&splashAfter=true" target="_blank">Demo: Interactive visualization of 80,000+ Facebook friendships</a> (<a href="http://snap.stanford.edu" target="_blank">source data</a></em>) </td> </tr> </table>

PyGraphistry is an open source Python library for data scientists and developers to leverage the power of graph visualization, analytics, AI, including with native GPU acceleration:

Python dataframe-native graph processing: Quickly ingest & prepare data in many formats, shapes, and scales as graphs. Use tools like Pandas, Spark, RAPIDS (GPU), and Apache Arrow.
Integrations: Connect to graph databases, data platforms, Python tools, and more.

Category Connector Tutorials
Data Platforms, SQL & Logs
Graph Databases
Python Tools & Libraries

View all connectors →
Prototype locally and deploy remotely: Prototype from notebooks like Jupyter and Databricks using local CPUs & GPUs, and then power production dashboards & pipelines with Graphistry Hub and your own self-hosted servers.
Query graphs with GFQL: Use GFQL, the first fully vectorized dataframe-native graph query language with an open-source GPU runtime, to ask relationship questions that are difficult for tabular tools without requiring a database. It supports friendly Cypher syntax and declarative graph semantics through g.gfql("MATCH ..."), with the same execution model available on the current bound graph or remotely via g.gfql_remote([...]).
graphistry[ai]: Call streamlined graph ML & AI methods to benefit from clustering, UMAP embeddings, graph neural networks, automatic feature engineering, and more.
Visualize & explore large graphs: In just a few minutes, create stunning interactive visualizations with millions of edges and many point-and-click built-ins like drilldowns, timebars, and filtering. When ready, customize with Python, JavaScript, and REST APIs.
Columnar & GPU acceleration: CPU-mode ingestion and wrangling is fast due to native use of Apache Arrow and columnar analytics, and the optional RAPIDS-based GPU mode delivers 100X+ speedups.

Category	Connector Tutorials
Data Platforms, SQL & Logs
Graph Databases
Python Tools & Libraries

From global 10 banks, manufacturers, news agencies, and government agencies, to startups, game companies, scientists, biotechs, and NGOs, many teams are tackling their graph workloads with Graphistry.

AI Assistant Integration

For LLM coding assistants (Claude Code, Cursor, Codex, etc.), install the official graphistry-skills package for better PyGraphistry code generation:

bash
npx skills add graphistry/graphistry-skills

Skills improve AI success rates from ~50% to ~90% on PyGraphistry tasks by providing context-aware guidance for graph ETL, visualization, GFQL queries, and AI workflows.

Gallery

The notebook demo gallery shares many more live visualizations, demos, and integration examples

<table> <tr valign="top"> <td width="33%" align="center"><a href="https://hub.graphistry.com/graph/graph.html?dataset=Twitter&splashAfter=true" target="_blank">Twitter Botnet<br><img width="266" src="https://i.imgur.com/qm5MCqS.jpg"></a></td> <td width="33%" align="center">Edit Wars on Wikipedia<br><a href="https://i.imgur.com/074zFve.png" target="_blank"><img width="266" src="https://i.imgur.com/074zFve.png"></a><em>(<a href="https://snap.stanford.edu" target="_blank">data</a></em>)</td> <td width="33%" align="center"><a href="https://hub.graphistry.com/graph/graph.html?dataset=bitC&splashAfter=true" target="_blank">100,000 Bitcoin Transactions<br><img width="266" height="266" src="https://i.imgur.com/axIkjfd.png"></a></td> </tr> <tr valign="top"> <td width="33%" align="center">Port Scan Attack<br><a href="http://i.imgur.com/vKUDySw.png" target="_blank"><img width="266" src="http://i.imgur.com/vKUDySw.png"></a></td> <td width="33%" align="center"><a href="http://hub.graphistry.com/graph/graph.html?dataset=PyGraphistry/M9RL4PQFSF&usertag=github&info=true&static=true&contentKey=Biogrid_Github_Demo&play=3000&center=false&menu=true&goLive=false&left=-2.58e+4&right=4.35e+4&top=-1.72e+4&bottom=2.16e+4&legend={%22title%22:%22%3Ch3%3EBioGRID%20Repository%20of%20Protein%20Interactions%3C/h3%3E%22,%22subtitle%22:%22%3Cp%3EEach%20color%20represents%20an%20organism.%20Humans%20are%20in%20light%20blue.%3C/p%3E%22,%22nodes%22:%22Proteins/Genes%22,%22edges%22:%22Interactions%20reported%20in%20scientific%20publications%22}" target="_blank">Protein Interactions <br><img width="266" src="http://i.imgur.com/nrUHLFz.png" target="_blank"></a><em>(<a href="http://thebiogrid.org" target="_blank">data</a>)</em></td> <td width="33%" align="center"><a href="http://hub.graphistry.com/graph/graph.html?&dataset=PyGraphistry/PC7D53HHS5&info=true&static=true&contentKey=SocioPlt_Github_Demo&play=3000&center=false&menu=true&goLive=false&left=-236&right=265&top=-145&bottom=134&usertag=github&legend=%7B%22nodes%22%3A%20%22%3Cspan%20style%3D%5C%22color%3A%23a6cee3%3B%5C%22%3ELanguages%3C/span%3E%20/%20%3Cspan%20style%3D%5C%22color%3Argb%28106%2C%2061%2C%20154%29%3B%5C%22%3EStatements%3C/span%3E%22%2C%20%22edges%22%3A%20%22Strong%20Correlations%22%2C%20%22subtitle%22%3A%20%22%3Cp%3EFor%20more%20information%2C%20check%20out%20the%20%3Ca%20target%3D%5C%22_blank%5C%22%20href%3D%5C%22https%3A//lmeyerov.github.io/projects/socioplt/viz/index.html%5C%22%3ESocio-PLT%3C/a%3E%20project.%20Make%20your%20own%20visualizations%20with%20%3Ca%20target%3D%5C%22_blank%5C%22%20href%3D%5C%22https%3A//github.com/graphistry/pygraphistry%5C%22%3EPyGraphistry%3C/a%3E.%3C/p%3E%22%2C%20%22title%22%3A%20%22%3Ch3%3ECorrelation%20Between%20Statements%20about%20Programming%20Languages%3C/h3%3E%22%7D" target="_blank">Programming Languages<br><img width="266" src="http://i.imgur.com/0T0EKmD.png"></a><em>(<a href="http://lmeyerov.github.io/projects/socioplt/viz/index.html" target="_blank">data</a>)</em></td> </tr> </table>

Install

Common configurations:

Minimal core

Includes: The GFQL dataframe-native graph query language, built-in layouts, Graphistry visualization server client
```
python
pip install graphistry
```
Does not include graphistry[ai], plugins

No dependencies and user-level

python
pip install --no-deps --user graphistry

GPU acceleration - Optional

Local GPU: Install RAPIDS and/or deploy a GPU-ready Graphistry server

Remote GPU: Use the remote endpoints.

For further options, see the installation guides

Visualization quickstart

Quickly go from raw data to a styled and interactive Graphistry graph visualization:

python
import graphistry
import pandas as pd

# Raw data as Pandas CPU dataframes, cuDF GPU dataframes, Spark, ...
df = pd.DataFrame({
    'src': ['Alice', 'Bob', 'Carol'],
    'dst': ['Bob', 'Carol', 'Alice'],
    'friendship': [0.3, 0.95, 0.8]
})

# Bind
g1 = graphistry.edges(df, 'src', 'dst')

# Override styling defaults
g1_styled = g1.encode_edge_color('friendship', ['blue', 'red'], as_continuous=True)

# Connect: Free GPU accounts and self-hosting @ graphistry.com/get-started
graphistry.register(api=3, username='your_username', password='your_password')

# Upload for GPU server visualization session
g1_styled.plot()

Explore 10 Minutes to Graphistry Visualization for more visualization examples and options

PyGraphistry[AI] & GFQL quickstart - CPU & GPU

CPU graph pipeline combining graph ML, AI, mining, and visualization:

python
from graphistry import n, e, e_forward, e_reverse

# Graph analytics
g2 = g1.compute_igraph('pagerank')
assert 'pagerank' in g2._nodes.columns

# Graph ML/AI
g3 = g2.umap()
assert ('x' in g3._nodes.columns) and ('y' in g3._nodes.columns)

# Graph querying with GFQL
g4 = g3.gfql([
    n(query='pagerank > 0.1'), e_forward(), n(query='pagerank > 0.1')
])
assert (g4._nodes.pagerank > 0.1).all()

# Upload for GPU server visualization session
g4.plot()

The automatic GPU modes require almost no code changes:

python
import cudf
from graphistry import n, e, e_forward, e_reverse

# Modified -- Rebind data as a GPU dataframe and swap in a GPU plugin call
g1_gpu = g1.edges(cudf.from_pandas(df))
g2 = g1_gpu.compute_cugraph('pagerank')

# Unmodified -- Automatic GPU mode for all ML, AI, GFQL queries, & visualization APIs
g3 = g2.umap()
g4 = g3.gfql([
    n(query='pagerank > 0.1'), e_forward(), n(query='pagerank > 0.1')
])
g4.plot()

Explore 10 Minutes to PyGraphistry for a wider variety of graph processing.

PyGraphistry documentation

Main PyGraphistry documentation
10 Minutes to: PyGraphistry, Visualization, GFQL
Get started: Install, UI Guide, Notebooks
Performance: PyGraphistry CPU+GPU & GFQL CPU+GPU
API References
- PyGraphistry API Reference: Visualization & Compute, PyGraphistry Cheatsheet
- GFQL Documentation: GFQL Cheatsheet and GFQL Operator Cheatsheet
- Plugins: Databricks, Splunk, Neptune, Neo4j, RAPIDS, and more
- Web: iframe, JavaScript, REST

Graphistry ecosystem

Graphistry server:
- Launch - Graphistry Hub, Graphistry cloud marketplaces, and self-hosting
- Self-hosting: Administration (including Docker) & Kubernetes
Graphistry client APIs:
- Web: iframe, JavaScript, REST
- PyGraphistry
- Graphistry for Microsoft PowerBI
Additional projects:
- Louie.ai: GenAI-native notebooks & dashboards to talk to your databases & Graphistry
- graph-app-kit: Streamlit Python dashboards with batteries-include graph packages
- cu-cat: Automatic GPU feature engineering