Quick answer
json.dumps() was handed a value whose type has no JSON equivalent — a datetime, Decimal, set, bytes, UUID, or a custom object. The quickest fix is json.dumps(obj, default=str); for control, convert each type yourself in a default function or a custom JSONEncoder.
The exact error string
The traceback ends with the type that could not be converted — datetime is the most common, but the same error names Decimal, set, bytes, UUID, and your own classes:
Traceback (most recent call last):
File "app.py", line 5, in <module>
json.dumps(data)
TypeError: Object of type datetime is not JSON serializableWhy it happens
The json module maps a fixed set of Python types to JSON: dict → object, list/tuple → array, str → string, int/float → number, True/False → true/false, and None → null. Anything else has no JSON representation, so json.dumps() stops and raises TypeError rather than guess how to encode it.
Cause: datetime / date
import json
from datetime import datetime
data = {"created": datetime.now()}
json.dumps(data)
# TypeError: Object of type datetime is not JSON serializableConvert the datetime to an ISO-8601 string. The quickest one-liner uses default=str; for explicit control use isoformat() in a default function:
import json
from datetime import datetime, date
data = {"created": datetime.now()}
# Quick fix — stringify anything json can't handle
json.dumps(data, default=str)
# {"created": "2026-06-18 14:03:55.123456"}
# Explicit fix — control the format
def encode(obj):
if isinstance(obj, (datetime, date)):
return obj.isoformat()
raise TypeError(f"Object of type {type(obj).__name__} is not JSON serializable")
json.dumps(data, default=encode)
# {"created": "2026-06-18T14:03:55.123456"}Cause: Decimal
import json
from decimal import Decimal
json.dumps({"price": Decimal("19.99")})
# TypeError: Object of type Decimal is not JSON serializable
# Fix: float() (small precision risk) or str() (exact)
json.dumps({"price": Decimal("19.99")}, default=float) # {"price": 19.99}
json.dumps({"price": Decimal("19.99")}, default=str) # {"price": "19.99"}Cause: set, bytes, UUID
import json
from uuid import uuid4
def encode(obj):
if isinstance(obj, set):
return list(obj) # JSON has no set type
if isinstance(obj, bytes):
return obj.decode("utf-8") # or base64 for binary data
return str(obj) # UUID and anything else
json.dumps({"tags": {"a", "b"}, "id": uuid4()}, default=encode)Cause: NumPy types (np.int64, np.ndarray)
NumPy scalars and arrays are their own types — np.int64 is not a Python int, and np.ndarray is not a list — so json.dumps() rejects them. Convert scalars with .item() (or int()/float()) and arrays with .tolist():
import json
import numpy as np
data = {"count": np.int64(42), "values": np.array([1, 2, 3])}
json.dumps(data)
# TypeError: Object of type int64 is not JSON serializable
def encode(obj):
if isinstance(obj, np.integer):
return int(obj)
if isinstance(obj, np.floating):
return float(obj)
if isinstance(obj, np.ndarray):
return obj.tolist()
raise TypeError(f"Object of type {type(obj).__name__} is not JSON serializable")
json.dumps(data, default=encode)
# {"count": 42, "values": [1, 2, 3]}Cause: a custom class, dataclass, or Pydantic model
import json
from dataclasses import dataclass, asdict
@dataclass
class User:
name: str
age: int
user = User("Alice", 30)
json.dumps(user) # TypeError: Object of type User is not JSON serializable
json.dumps(asdict(user)) # {"name": "Alice", "age": 30}
# Plain class without dataclass:
json.dumps(user, default=lambda o: o.__dict__)
# Pydantic v2 model:
# json.dumps(model.model_dump()) or model.model_dump_json()The reusable fix: a custom JSONEncoder
When the same types appear across a codebase, subclass json.JSONEncoder once and pass it with cls= everywhere:
import json
from datetime import datetime, date
from decimal import Decimal
from uuid import UUID
class SafeEncoder(json.JSONEncoder):
def default(self, obj):
if isinstance(obj, (datetime, date)):
return obj.isoformat()
if isinstance(obj, Decimal):
return float(obj)
if isinstance(obj, (set, frozenset)):
return list(obj)
if isinstance(obj, (bytes, UUID)):
return str(obj)
return super().default(obj)
json.dumps(data, cls=SafeEncoder)Check your serialized output
Once json.dumps() succeeds, paste the result into the formatter to inspect and confirm it is valid JSON. Everything runs in your browser; nothing is uploaded.
Frequently Asked Questions
What does TypeError: Object of type datetime is not JSON serializable mean?
It means json.dumps() received a value whose Python type has no built-in JSON representation. The json module only knows how to convert dict, list, tuple, str, int, float, bool, and None. Types like datetime, Decimal, set, bytes, UUID, and custom classes are not in that list, so it raises TypeError unless you tell it how to convert them.
How do I fix Object of type datetime is not JSON serializable?
Pass a default function to json.dumps() that converts the unsupported type. For dates, json.dumps(obj, default=str) turns datetime into a string. For full control, call obj.isoformat() inside a custom default function, or subclass json.JSONEncoder and override its default() method.
Why does Decimal raise not JSON serializable?
decimal.Decimal is a separate numeric type that json does not treat as a number, so it raises TypeError. Convert it with float(value) if minor precision loss is acceptable, or str(value) to preserve the exact value. A default such as default=lambda o: float(o) handles every Decimal in one place.
How do I serialize a Python set to JSON?
JSON has no set type, so json.dumps() raises TypeError on a set. Convert it to a list first — json.dumps(list(my_set)) — or handle it in a default function. Note that converting to a list does not guarantee order, and the value deserializes back as a list, not a set.
How do I make a custom class JSON serializable?
Give json.dumps() a way to turn the object into JSON-compatible types. The simplest options are a default function returning vars(obj) or obj.__dict__, a dataclass with dataclasses.asdict(obj), or a Pydantic model with model.model_dump(). For reusable behavior, subclass json.JSONEncoder and implement default().