encode/httpx · error · TypeError

Unexpected type for 'content', {type(content)!r}

Error message

Unexpected type for 'content', {type(content)!r}

What it means

This TypeError is the terminal fallthrough of httpx._content.encode_content. The function accepts str, bytes, an Iterable[bytes] (non-dict), or an AsyncIterable[bytes]; anything else reaching line 133 is rejected. The dict-specific guard a few lines up exists because dict is iterable but almost always a caller mistake (issue #2491).

Source

Thrown at httpx/_content.py:133

    elif isinstance(content, Iterable) and not isinstance(content, dict):
        # `not isinstance(content, dict)` is a bit oddly specific, but it
        # catches a case that's easy for users to make in error, and would
        # otherwise pass through here, like any other bytes-iterable,
        # because `dict` happens to be iterable. See issue #2491.
        content_length_or_none = peek_filelike_length(content)

        if content_length_or_none is None:
            headers = {"Transfer-Encoding": "chunked"}
        else:
            headers = {"Content-Length": str(content_length_or_none)}
        return headers, IteratorByteStream(content)  # type: ignore

    elif isinstance(content, AsyncIterable):
        headers = {"Transfer-Encoding": "chunked"}
        return headers, AsyncIteratorByteStream(content)

    raise TypeError(f"Unexpected type for 'content', {type(content)!r}")


def encode_urlencoded_data(
    data: RequestData,
) -> tuple[dict[str, str], ByteStream]:
    plain_data = []
    for key, value in data.items():
        if isinstance(value, (list, tuple)):
            plain_data.extend([(key, primitive_value_to_str(item)) for item in value])
        else:
            plain_data.append((key, primitive_value_to_str(value)))
    body = urlencode(plain_data, doseq=True).encode("utf-8")
    content_length = str(len(body))
    content_type = "application/x-www-form-urlencoded"
    headers = {"Content-Length": content_length, "Content-Type": content_type}
    return headers, ByteStream(body)

View on GitHub (pinned to b5addb64f0)

Solutions

  1. Use json=<value> for dict/structured data instead of content=<dict>.
  2. Coerce to bytes before passing: content=str(value).encode() or content=bytes(value).
  3. For a list of strings, map to bytes: content=[s.encode() for s in items].
  4. If you have an arbitrary object, serialize it first (json.dumps(...).encode()).

Example fix

// before
client.post(url, content={'key': 'val'})  # TypeError
// after
client.post(url, json={'key': 'val'})
// or
client.post(url, content=json.dumps({'key':'val'}).encode())
Defensive patterns

Strategy: type-guard

Validate before calling

def normalize_content(content):
    if isinstance(content, (bytes, str)):
        return content
    if isinstance(content, dict):
        raise TypeError('dict content should use json= instead of content=')
    if isinstance(content, (list, tuple)):
        return [c.encode() if isinstance(c, str) else c for c in content]
    raise TypeError(f'content must be str/bytes/iterable-of-bytes, got {type(content)!r}')

Type guard

import collections.abc as cabc

def is_valid_content(content) -> bool:
    if isinstance(content, (bytes, str)):
        return True
    if isinstance(content, dict):
        return False
    if isinstance(content, cabc.Iterable):
        return all(isinstance(x, bytes) for x in content)
    return False

Try / catch

try:
    resp = client.post(url, content=payload)
except TypeError as exc:
    if 'Unexpected type' in str(exc):
        resp = client.post(url, json=payload)
    else:
        raise

Prevention

When it happens

Trigger: Passing content=<dict> with a dict that slipped past the guard (or a dict subclass); passing content=<int>/<float>/<None substitute>/custom object; calling Response(content=...) or Request(content=...) with a non-byte-iterable like a list of str; passing a non-iterable object such as a number.

Common situations: Passing content=some_dict instead of json=some_dict; passing content=123; passing content=["a","b"] (list of str, not bytes); passing content=None-equivalent or a numpy array / pandas object without conversion; mixing up content= and data= semantics.

Related errors


AI-assisted analysis of encode/httpx@b5addb64f0 (2026-08-04). Data as JSON: /data/errors/d2db72028205eacd.json. Report an issue: GitHub.