cURL to Python requests Converter

curl commands into Python scripts

Python is where many API calls end up: automation scripts, data pipelines, notebooks and tests. Starting from a curl command is common — copied from API documentation, from Postman, or from the browser’s network panel with “Copy as cURL”. This converter writes requests code that sends the same bytes as curl, using the library’s keyword arguments wherever that is safe, so the result reads like code a Python developer would write by hand.

Keyword arguments generated

• Method: requests.get, .post, .put, .patch, .delete, .head or .options; other verbs use requests.request("PURGE", ...).
• params: a simple query string is split into a params dict. If any part needs percent-encoding or a name repeats, the URL is kept as written.
• headers: a dict of your -H values; duplicates are merged.
• cookies: a -b 'session=abc; theme=dark' cookie string becomes a cookies dict when every pair is a plain token.
• json: a JSON object or array declared as JSON is written as a Python literal (True, False, None) and passed as json=json_data; the redundant Content-Type that curl implied is dropped because requests sets it.
• data: url-encoded bodies become a dict when requests would encode it back to exactly the same bytes; anything else is passed as the original string. Non-ASCII text is sent as UTF-8 via .encode(), matching curl.
• files: -F fields become a files mapping, with open(path, "rb") for uploads, a file name and an optional content type; text fields use (None, value). Repeated field names switch to a list of tuples.
• auth: -u user:pass becomes auth=("user", "pass").
• timeout: -m 30 gives timeout=30; adding --connect-timeout 5 gives timeout=(5, 30).
• verify=False for -k, and proxies for -x, with -U credentials embedded in the proxy URL.

Redirects: a classic gotcha

requests follows redirects automatically for every method except HEAD, but curl only does so with -L. The generated code therefore adds allow_redirects=False when the command lacks -L, and allow_redirects=True for a HEAD request that has it. Delete the argument if you actually want the redirect followed. --compressed needs no code, since requests always accepts gzip and deflate; an info note confirms that.

Output style and what is not converted

The code is laid out the way Black and Ruff format it — double quotes, one argument per line with trailing commas once a call gets long — so running a formatter changes nothing. It prints response.status_code and response.text (the headers for HEAD).

Options that save output or tune curl itself (-o, --retry, -w) are reported as not converted. --digest, --ntlm and --negotiate fall back to Basic auth with a warning; requests needs HTTPDigestAuth or an extra package for those. As with every tool here, nothing you paste is uploaded. To clean up the rest of your script, try the Python formatter.

Examples

curl -X POST 'https://api.example.com/v1/orders?expand=items' \
  -H 'Content-Type: application/json' \
  -b 'session=abc123; theme=dark' \
  --connect-timeout 5 -m 30 \
  -d '{"customer":"cus_42","gift":true,"note":null}'

import requests

params = {
    "expand": "items",
}

headers = {
    "Content-Type": "application/json",
}

cookies = {
    "session": "abc123",
    "theme": "dark",
}

json_data = {
    "customer": "cus_42",
    "gift": True,
    "note": None,
}

response = requests.post(
    "https://api.example.com/v1/orders",
    params=params,
    headers=headers,
    cookies=cookies,
    json=json_data,
    timeout=(5, 30),
    allow_redirects=False,
)

print(response.status_code)
print(response.text)

curl https://files.example.com/upload \
  -u reports:Tr0ub4dor \
  -F '[email protected];type=application/pdf' \
  -F 'folder=finance'

import requests

files = {
    "file": ("q3-report.pdf", open("q3-report.pdf", "rb"), "application/pdf"),
    "folder": (None, "finance"),
}

response = requests.post(
    "https://files.example.com/upload",
    auth=("reports", "Tr0ub4dor"),
    files=files,
    allow_redirects=False,
)

print(response.status_code)
print(response.text)

curl -L https://shop.example.com/cart -d 'sku=KB-104' -d 'qty=2' -d 'note=gift+wrap'

import requests

data = {
    "sku": "KB-104",
    "qty": "2",
    "note": "gift wrap",
}

response = requests.post(
    "https://shop.example.com/cart",
    data=data,
)

print(response.status_code)
print(response.text)

Common errors and how to fix them

Related tools

• cURL to JavaScript fetch
• cURL to Axios
• cURL to Go net/http
• cURL to PHP
• Python formatter
• JSON formatter