curl-cffi-fingerprints 0.16.0.151.2

Drop-in curl_cffi fork with automatically harvested browser fingerprints for new Chrome and Safari releases.

curl-cffi-fingerprints

A drop-in fork of curl_cffi whose one job is keeping browser fingerprints fresh: every new Chrome and Safari release gets its impersonation profile harvested, verified, and shipped automatically.

curl_cffi is a Python binding for curl-impersonate via cffi. Unlike pure-Python HTTP clients such as httpx or requests, it can impersonate browsers' TLS/JA3 and HTTP/2 fingerprints, so requests look like they come from a real browser.

This fork changes nothing about the API. It exists because upstream only adds impersonation targets when there is a protocol-level change worth supporting, while scraping in practice often needs the current browser version. Here, new versions land as soon as they are released.

How it works

The fingerprints are harvested and shipped by an automated pipeline across two repositories:

1. aleixrodriala/curl-impersonate captures profiles from real browsers for each new release — Chrome (desktop and Android) and Safari (macOS and iOS) — as declarative JSON profiles, and publishes a libcurl-impersonate build.
2. That release triggers a sync workflow in this repository, which adds the new targets, bumps the version, builds wheels for all supported platforms, and publishes them to PyPI and GitHub Releases.

No manual step in between: a new browser release becomes a new installable version of this package.

Versioning

Versions follow <upstream-base>.<chrome-major>.<revision>. For example, 0.16.0.151.2 is upstream curl_cffi 0.16.0 plus the fingerprint drop for Chrome 151, revision 2. Upgrading within the same upstream base only adds fingerprints; API changes come from tracking upstream releases.

Install

pip install curl-cffi-fingerprints --upgrade

This works on Linux, macOS and Windows out of the box.

The import name is still curl_cffi — existing code keeps working unchanged. Because it provides the same import package, uninstall the upstream package first if you have it:

pip uninstall -y curl-cffi
pip install curl-cffi-fingerprints --upgrade

Wheels are also attached to GitHub Releases if you prefer pinning an exact fingerprint drop.

Usage

The API is identical to upstream curl_cffi: a requests-like high-level API, a low-level curl API, asyncio, WebSockets, and the curl-cffi CLI. Full documentation lives at curl-cffi.readthedocs.io.

import curl_cffi

# Notice the impersonate parameter: "chrome" always resolves to the newest
# harvested Chrome fingerprint, so you stay current just by upgrading.
r = curl_cffi.get("https://tls.browserleaks.com/json", impersonate="chrome")
print(r.json())

# Similar rolling aliases: "chrome_android", "safari", "safari_ios"
r = curl_cffi.get("https://tls.browserleaks.com/json", impersonate="safari_ios")

# Or pin a specific version
r = curl_cffi.get("https://tls.browserleaks.com/json", impersonate="chrome151")

# http/3 with impersonation
r = curl_cffi.get("https://example.com", http_version="v3", impersonate="chrome")

# Bring your own fingerprints for non-browser targets
r = curl_cffi.get("https://tls.browserleaks.com/json", ja3=..., akamai=...)

Sessions and asyncio:

s = curl_cffi.Session()
s.get("https://httpbin.org/cookies/set/foo/bar")
print(s.cookies)

from curl_cffi import AsyncSession

async with AsyncSession() as s:
    r = await s.get("https://example.com")

CLI:

# Debug a URL with impersonation
curl-cffi get tls.browserleaks.com/json --impersonate chrome

# List the fingerprints available on your install
curl-cffi list

Supported targets

All upstream targets are included, plus the fork's harvested profiles for new Chrome (desktop and Android) and Safari (macOS and iOS) releases. To see exactly what your installed version supports, run curl-cffi list; the profile sources live in the curl-impersonate fork.

Relationship to upstream

This fork stays intentionally thin: all library development happens in lexiforest/curl_cffi, which this fork tracks and merges regularly. Only the fingerprint pipeline and the harvested profiles are added here.

• Library bugs or feature requests → report them upstream.
• Fingerprint issues (wrong/missing profile for a browser version) → report them at the curl-impersonate fork.

If you need commercial support or fingerprints beyond Chrome/Safari, check the upstream ecosystem at impersonate.pro.

Acknowledgement

• Forked from lexiforest/curl_cffi by @lexiforest, which was originally forked from multippt/python_curl_cffi. Both are under the MIT license.
• Built on curl-impersonate by @lwthiker and the lexiforest fork.
• Headers/Cookies files are copied from httpx, which is under the BSD license.

MIT licensed.