lithops
Lithops lets you transparently run your Python applications in the Cloud
Decision gist · record as of 2026-08-14
Yes, if you have highly parallel workloads and want to avoid cloud-specific code. The framework is actively maintained, has no known vulnerabilities, and supports a wide range of backends. Install friction is low. The main gotcha is that you must configure a compute and storage backend before running anything; it is not a drop-in replacement for local parallelism without setup. Best for teams already using cloud or Kubernetes and wanting code portability across providers.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires configuration of at least one compute backend (e.g., AWS Lambda, Google Cloud Run, Kubernetes) and corresponding storage backend before execution; localhost backends available for testing.
- Low friction installation with a pure-Python wheel.
- Active maintenance with recent releases; last commit 2026-07-17.
License · maintenance · safety
permissive license (permissive) — Licensed under Apache Software License (permissive). No restrictions on commercial or proprietary use; you may modify and redistribute under the same license terms.
last release 2026-07-02 (43 days) · last repo commit 2026-07-17 · 367 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 233,113 downloads/mo, #9,045 on PyPI
Alternatives
Verify before relying
pip install lithops
from lithops import FunctionExecutor
def double(i):
return i * 2
with FunctionExecutor() as fexec:
f = fexec.map(double, [1, 2, 3, 4])
print(f.result())- Whether the 12 runtime dependencies introduce any transitive security concerns beyond the zero known vulnerabilities in lithops itself.
- Performance characteristics and cost implications when switching between backends (e.g., Lambda vs. Batch vs. on-premise).
- Maturity and stability of less-common backends (Aliyun, Oracle, OpenWhisk, Singularity) relative to AWS/GCP/Azure.
What it is and what it does
Lithops abstracts away cloud infrastructure management by letting you write ordinary Python functions and run them in parallel across multiple cloud providers, Kubernetes clusters, or on-premise systems without code changes. It supports both a Futures-style API (map/reduce patterns) and a Multiprocessing API (drop-in replacement for Python's multiprocessing.Pool), plus cloud-native storage backends for reading and writing data. The framework handles deployment, scaling, and orchestration transparently—you configure a compute backend (Lambda, Cloud Run, Batch, Kubernetes, etc.) and a storage backend (S3, GCS, Azure Blob, etc.), then submit your workload.
The primary use case is highly parallel workloads: Monte Carlo simulations, machine learning preprocessing, geospatial analytics, and metabolomics. You can optimize for speed using serverless functions that launch hundreds of tasks in milliseconds, or for cost by running the same code on batch systems with spot instances. It supports major cloud providers (AWS, Google Cloud, Azure, IBM, Aliyun, Oracle) and on-premise platforms (Kubernetes, OpenStack, Ceph, MinIO), making it useful for hybrid or multi-cloud scenarios where code portability matters.
Use it for
- Run Monte Carlo simulations or financial modeling across hundreds of parallel functions on AWS Lambda without managing infrastructure.
- Process large geospatial or metabolomics datasets by distributing computation across Kubernetes pods or cloud VMs.
- Migrate existing Python multiprocessing code to serverless or cloud backends by swapping Pool() with lithops.multiprocessing.Pool().
- Execute the same data pipeline on multiple cloud providers (AWS, GCP, Azure) by changing configuration without rewriting code.
- Reduce costs for batch workloads by running them on Spot Instances via AWS Batch while keeping the same Python code.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you have highly parallel workloads and want to avoid cloud-specific code.
The framework is actively maintained, has no known vulnerabilities, and supports a wide range of backends. Install friction is low. The main gotcha is that you must configure a compute and storage backend before running anything; it is not a drop-in replacement for local parallelism without setup. Best for teams already using cloud or Kubernetes and wanting code portability across providers.
Install
lithops on PyPI
Before you install
Low friction installation with a pure-Python wheel. Active maintenance with recent releases; last commit 2026-07-17. Depends on 12 runtime packages including Click, PyYAML, cloudpickle, paramiko, and pika—all standard libraries for CLI, configuration, serialization, SSH, and messaging.
Requires configuration of at least one compute backend (e.g., AWS Lambda, Google Cloud Run, Kubernetes) and corresponding storage backend before execution; localhost backends available for testing.
License in practice
Licensed under Apache Software License (permissive). No restrictions on commercial or proprietary use; you may modify and redistribute under the same license terms.
Quickstart
pip install lithops
from lithops import FunctionExecutor
def double(i):
return i * 2
with FunctionExecutor() as fexec:
f = fexec.map(double, [1, 2, 3, 4])
print(f.result())Verify before relying
- Whether the 12 runtime dependencies introduce any transitive security concerns beyond the zero known vulnerabilities in lithops itself.
- Performance characteristics and cost implications when switching between backends (e.g., Lambda vs. Batch vs. on-premise).
- Maturity and stability of less-common backends (Aliyun, Oracle, OpenWhisk, Singularity) relative to AWS/GCP/Azure.
Package facts
| License | permissive license permissive |
| Python support | Supports the current Python release >=3.6 |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 12 packagesClickPyYAMLcloudpickleparamikopikaps-mempsutilrequestssixtabulatetblibtqdm |
| Maintenance | Actively maintained 43 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 233,113 / month, #9,045 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 5 - Production/StableIntended Audience :: DevelopersLicense :: OSI Approved :: Apache Software LicenseNatural Language :: EnglishOperating System :: OS IndependentProgramming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14Topic :: Scientific/EngineeringTopic :: Software Development :: Libraries :: Python ModulesTopic :: System :: Distributed Computing |
Evidence: lithops-3.7.0-py3-none-any.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “serverless multi-cloud execution”
- lithopsLithops is a Python framework for running unmodified code at scale…
- c7nCloud Custodian is a rules engine that enforces cloud infrastructure…
- deeplakeDeep Lake is a serverless database for storing, searching, and…
Give your agent the search over MCP, or paste the wish link into any chat.
More Scientific/Engineering packages
NumPy provides an N-dimensional array object and a comprehensive suite of mathematical, linear algebra, Fourier transform, and random number functions for scientific computing in Python.
pandas provides fast, flexible data structures (Series and DataFrame) for loading, cleaning, transforming, and analyzing labeled or relational data in Python.
scipy provides numerical algorithms for mathematics, science, and engineering—including optimization, integration, linear algebra, Fourier transforms, signal and image processing, and ODE solvers—built on numpy arrays.
scikit-learn provides a comprehensive Python library for supervised and unsupervised machine learning, including classification, regression, clustering, dimensionality reduction, and model evaluation tools built on NumPy and SciPy.
Install it if you need to train, evaluate, or deploy supervised or unsupervised learning models.
dill extends Python's pickle module to serialize and deserialize a much wider range of Python objects, including functions, lambdas, classes, and interpreter sessions, to byte streams for storage or network transmission.
Multiprocess is an enhanced fork of Python's standard multiprocessing library that uses dill for better serialization, allowing you to spawn processes with a threading-like API and share complex objects between them.
Install it if you use multiprocessing and encounter pickle serialization limits with lambdas or complex objects.
See also skypilot-nightly · modal · distributed · aws-parallelcluster · polars-cloud · argo-workflows · litdata · torchx · fairscale · aiomultiprocess