--- title: uv vs conda: A Friendly Comparison From a Longtime conda... description: A longtime conda user compares uv and conda across speed, reproducibility, and use cases in 2026. uv wins for pure Python projects thanks to its Rust-based... canonical: https://daily.dev/posts/uv-vs-conda-a-friendly-comparison-from-a-longtime-conda-user-rhxklrdsm twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: uv vs conda: A Friendly Comparison From a Longtime conda User | daily.dev og:description: A longtime conda user compares uv and conda across speed, reproducibility, and use cases in 2026. uv wins for pure Python projects thanks to its Rust-based... og:url: https://daily.dev/posts/uv-vs-conda-a-friendly-comparison-from-a-longtime-conda-user-rhxklrdsm og:image: https://api.daily.dev/og/posts/RhxKlrdsM.png og:image:alt: uv vs conda: A Friendly Comparison From a Longtime conda User og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # uv vs conda: A Friendly Comparison From a Longtime conda User **[Medium](https://daily.dev/sources/medium_js)** · 9 min read · 1 upvotes · 0 comments ## Summary A longtime conda user compares uv and conda across speed, reproducibility, and use cases in 2026. uv wins for pure Python projects thanks to its Rust-based speed, built-in lockfile (uv.lock), and unified toolchain replacing pip, virtualenv, pyenv, and pip-tools. Conda remains the better choice when non-Python native dependencies are needed, such as the full GDAL stack, R interoperability, or exotic system libraries. The article recommends Miniforge over full Anaconda for licensing reasons, notes that PyTorch now favors pip wheels over conda, and introduces pixi as a hybrid tool that combines conda-forge packages with uv-speed and lockfile-first workflows. A practical decision table helps readers choose based on dependency type rather than community allegiance. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### Should I use uv or conda for my Python data science project? Choose based on dependency type: uv is best for pure Python projects, including most modern ML work with wheel-based PyTorch or JAX, because it is faster and reproducible via its uv.lock file. Conda, ideally via Miniforge, remains the better choice when dependencies include non-Python natives like the full GDAL stack, R, compilers, or other system libraries not available as wheels. _Compare uv and conda tradeoffs on daily.dev before locking in a dependency management workflow._ ### Do I still need conda for PyTorch or deep learning? Not necessarily. PyTorch stopped publishing to its official Anaconda channel after release 2.5 and now directs users toward pip wheels, with community builds surviving on conda-forge for those who want them. Modern GPU wheels bundle the CUDA runtime pieces they need, so running uv add torch covers most real-world deep learning setups despite the old folklore that deep learning requires conda. _Developers weighing PyTorch installation options can track ecosystem shifts like this on daily.dev._ ### How does uv achieve reproducible Python environments compared to conda's environment.yml? uv writes a uv.lock file by default that pins every transitive dependency with hashes and covers Linux, macOS, and Windows from the same file, so running uv sync rebuilds an identical environment anywhere. Conda's environment.yml is only a specification of requested packages, not a lock, so two conda env create runs weeks apart can produce different resolved versions unless the separate conda-lock tool is used. _Anyone chasing reproducible builds across machines can follow packaging tool comparisons like this on daily.dev._ ## Similar posts on daily.dev - [uv is the best thing to happen to the Python ecosystem in a decade](https://daily.dev/posts/uv-is-the-best-thing-to-happen-to-the-python-ecosystem-in-a-decade-14hjixazi) · Hacker News · 3 upvotes · 0 comments - [uv vs pip: Python Packaging and Dependency Management – Real Python](https://daily.dev/posts/uv-vs-pip-python-packaging-and-dependency-management-real-python-y2d0smxfy) · Real Python · 0 upvotes · 0 comments - [How to Manage Python Packages with uv](https://daily.dev/posts/how-to-manage-python-packages-with-uv-tblfaulqd) · freeCodeCamp · 14 upvotes · 0 comments - [Python Dependency Management: Is it Time to Drop Poetry for UV?](https://daily.dev/posts/python-dependency-management-is-it-time-to-drop-poetry-for-uv--dbig7y5fs) · Medium · 0 upvotes · 0 comments --- Tags: [#python](https://daily.dev/tags/python) [View this post on daily.dev](https://daily.dev/posts/uv-vs-conda-a-friendly-comparison-from-a-longtime-conda-user-rhxklrdsm) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"uv vs conda: A Friendly Comparison From a Longtime conda User","url":"https://daily.dev/posts/uv-vs-conda-a-friendly-comparison-from-a-longtime-conda-user-rhxklrdsm","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/uv-vs-conda-a-friendly-comparison-from-a-longtime-conda-user-rhxklrdsm"},"datePublished":"2026-07-23T12:09:51.250Z","dateModified":"2026-09-14T08:57:46.991Z","description":"A longtime conda user compares uv and conda across speed, reproducibility, and use cases in 2026. uv wins for pure Python projects thanks to its Rust-based...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/71d33fc8162eea2f3b4f0541b67c9122?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/71d33fc8162eea2f3b4f0541b67c9122?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Medium","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Medium","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/medium","url":"https://daily.dev/sources/medium_js"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/uv-vs-conda-a-friendly-comparison-from-a-longtime-conda-user-rhxklrdsm","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"python","timeRequired":"PT9M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Medium","item":"https://daily.dev/sources/medium_js"},{"@type":"ListItem","position":3,"name":"uv vs conda: A Friendly Comparison From a Longtime conda User"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/uv-vs-conda-a-friendly-comparison-from-a-longtime-conda-user-rhxklrdsm#faq","mainEntity":[{"@type":"Question","name":"Should I use uv or conda for my Python data science project?","acceptedAnswer":{"@type":"Answer","text":"Choose based on dependency type: uv is best for pure Python projects, including most modern ML work with wheel-based PyTorch or JAX, because it is faster and reproducible via its uv.lock file. Conda, ideally via Miniforge, remains the better choice when dependencies include non-Python natives like the full GDAL stack, R, compilers, or other system libraries not available as wheels. Compare uv and conda tradeoffs on daily.dev before locking in a dependency management workflow."}},{"@type":"Question","name":"Do I still need conda for PyTorch or deep learning?","acceptedAnswer":{"@type":"Answer","text":"Not necessarily. PyTorch stopped publishing to its official Anaconda channel after release 2.5 and now directs users toward pip wheels, with community builds surviving on conda-forge for those who want them. Modern GPU wheels bundle the CUDA runtime pieces they need, so running uv add torch covers most real-world deep learning setups despite the old folklore that deep learning requires conda. Developers weighing PyTorch installation options can track ecosystem shifts like this on daily.dev."}},{"@type":"Question","name":"How does uv achieve reproducible Python environments compared to conda's environment.yml?","acceptedAnswer":{"@type":"Answer","text":"uv writes a uv.lock file by default that pins every transitive dependency with hashes and covers Linux, macOS, and Windows from the same file, so running uv sync rebuilds an identical environment anywhere. Conda's environment.yml is only a specification of requested packages, not a lock, so two conda env create runs weeks apart can produce different resolved versions unless the separate conda-lock tool is used. Anyone chasing reproducible builds across machines can follow packaging tool comparisons like this on daily.dev."}}]} ```