SeleniumHQ/selenium · error · TypeError

`raw` must be a dict, got {type(raw)}

Error message

`raw` must be a dict, got {type(raw)}

What it means

The Proxy constructor accepts an optional `raw` argument that must be either None (use defaults) or a dict. Any other type raises TypeError naming the received type. The dict is then interpreted as a raw proxy-capability map (proxyType, httpProxy, noProxy, etc.).

Source

Thrown at py/selenium/webdriver/common/proxy.py:125

    socks_username = _ProxyTypeDescriptor("socksUsername", ProxyType.MANUAL)
    """SOCKS proxy username."""

    socks_password = _ProxyTypeDescriptor("socksPassword", ProxyType.MANUAL)
    """SOCKS proxy password."""

    socks_version = _ProxyTypeDescriptor("socksVersion", ProxyType.MANUAL)
    """SOCKS proxy version."""

    def __init__(self, raw: dict | None = None):
        """Creates a new Proxy.

        Args:
            raw: Raw proxy data. If None, default class values are used.
        """
        if raw is None:
            return
        if not isinstance(raw, dict):
            raise TypeError(f"`raw` must be a dict, got {type(raw)}")
        if raw.get("proxyType"):
            self.proxy_type = ProxyType.load(raw["proxyType"])
        if raw.get("httpProxy"):
            self.http_proxy = raw["httpProxy"]
        if raw.get("noProxy"):
            self.no_proxy = raw["noProxy"]
        if raw.get("proxyAutoconfigUrl"):
            self.proxy_autoconfig_url = raw["proxyAutoconfigUrl"]
        if raw.get("sslProxy"):
            self.sslProxy = raw["sslProxy"]
        if raw.get("autodetect"):
            self.auto_detect = raw["autodetect"]
        if raw.get("socksProxy"):
            self.socks_proxy = raw["socksProxy"]
        if raw.get("socksUsername"):
            self.socks_username = raw["socksUsername"]
        if raw.get("socksPassword"):
            self.socks_password = raw["socksPassword"]

View on GitHub (pinned to aa36b38e69)

Solutions

1. Pass a dict (or None): Proxy({'httpProxy': 'host:8080'}).
2. If you have a JSON string, json.loads() it first.
3. To copy a Proxy, pass proxy.to_capabilities().

Example fix

# before
p = Proxy('{"httpProxy": "host:8080"}')  # str -> TypeError

# after
import json
p = Proxy(json.loads('{"httpProxy": "host:8080"}'))
# or directly
p = Proxy({'httpProxy': 'host:8080'})

import json
if isinstance(raw, str):
    raw = json.loads(raw)
if raw is not None and not isinstance(raw, dict):
    raise TypeError('raw must be a dict or None')
proxy = Proxy(raw)

def is_proxy_raw(v) -> bool:
    return v is None or isinstance(v, dict)

try:
    proxy = Proxy(raw)
except TypeError:
    proxy = Proxy(raw if isinstance(raw, dict) else None)

When it happens

Trigger: Calling Proxy('MANUAL'), Proxy([('httpProxy','host')], Proxy(123), or Proxy(json_string). Passing a JSON-decoded value that is a list or scalar instead of an object.

Common situations: Passing a JSON string instead of a parsed dict. Passing a list of tuples. Passing another Proxy instance instead of its capabilities.

Related errors

• Only Proxy objects can be passed in.
• Autodetect proxy value needs to be a boolean
• no_proxy list must contain only strings
• no_proxy must be a comma-separated string or a list of strin
• Argument to isShown must be of type Element

AI-assisted analysis of SeleniumHQ/selenium@aa36b38e69 (2026-08-14). Data as JSON: /api/errors/7277852f6c0b6e5d. Report an issue: GitHub.