Loading page
About this tool
Build a pip requirements.txt with pinned versions, extras and environment markers, validated against PEP 508 and PEP 440 rules.
This generator builds a pip requirements.txt line by line, validating package names, extras and environment markers against PEP 508 and version specifiers against PEP 440 — including exact pins (==), compatible releases (~=), wildcards (==2.0.*) and markers like python_version < "3.11". It is for Python developers who want a correct, reproducible dependency file without memorising specifier grammar, and it warns about unpinned or unbounded requirements.
Open Requirements TXT Generator on AltFTool — it loads instantly in your browser.
Provide your input — an image, text, or data.
Let the tool analyze or generate the result.
Review, refine, and reuse the output wherever you need it.
Names, extras, versions and markers are checked against the actual PEP 508/440 grammar, not just non-empty.
One-click markers for Python version, Windows/macOS/Linux and arm64 — quoting handled for you.
Flags unpinned packages and >= ranges with no upper bound before they break a future install.
== pins one exact version (requests==2.31.0 installs only 2.31.0), while ~= allows compatible releases: ~=2.31.0 means >=2.31.0 but <2.32.0, and ~=2.2 means >=2.2 but <3.0. Applications usually pin with ==; libraries usually declare ranges.
Append an environment marker after a semicolon: tomli>=1.1.0 ; python_version < "3.11". The comparison value must be in quotes. PEP 508 markers can also test sys_platform, platform_machine and os_name, and combine clauses with and/or.
They install the package's optional extras — additional dependency groups the project defines. requests[socks] pulls in the SOCKS-proxy dependencies alongside requests itself; multiple extras are comma-separated, like uvicorn[standard] or celery[redis,msgpack].
For deployed applications, yes — exact pins (plus hashes via pip-compile --generate-hashes) are the only way to guarantee the same install twice. For libraries, no — pin ranges instead, because exact pins in a library force conflicts onto every consumer. A common setup is a loose pyproject.toml plus a fully pinned, generated requirements lock file.