Prisma Client Python

Warnings

• breaking The Prisma Client Python project has been officially abandoned by its maintainer as of April 15, 2025, and its GitHub repository is now read-only. No further development, bug fixes, or new features will be added. Users should consider alternative ORMs or be prepared to maintain a fork.
  Fix: Evaluate migration to a different ORM (e.g., SQLAlchemy, SQLModel) or be prepared to maintain a fork of this project.
• breaking Support for Python 3.7 was dropped in version 0.14.0, as Python 3.7 reached its End-of-Life.
  Fix: Upgrade your Python environment to 3.8 or newer.
• gotcha The Prisma Client Python is auto-generated based on your `schema.prisma` file. You must run `prisma generate` or `prisma db push` (which also generates the client) every time you modify your `schema.prisma` file for changes to be reflected in your Python code. Ensure your `generator client` block in `schema.prisma` specifies `provider = "prisma-client-py"`.
  Fix: Always execute `prisma db push` or `prisma generate` in your terminal after making changes to `schema.prisma`.
• gotcha When upgrading to Pydantic v2 (or using it alongside v1), there might be compatibility issues with `prisma.validate`. Specifically, errors raised by `prisma.validate` might need to be caught using `pydantic.v1.ValidationError` due to internal implementation details in v0.10.0.
  Fix: Import `ValidationError` from `pydantic.v1` for error handling related to `prisma.validate` if encountering issues with Pydantic v2.
• gotcha Prisma Client Python's static type checking has limitations, especially when using `mypy`. It is highly recommended to use `pyright` as your type checker for Prisma Client Python applications to achieve the most accurate and comprehensive type safety.
  Fix: Configure your development environment to use `pyright` as the type checker.

Install

• pip install prisma Install latest version

Imports

• Prisma

  from prisma import Prisma

Quickstart

To use Prisma Client Python, you must first define your database schema in a `schema.prisma` file and then run `prisma db push` (or `prisma generate`) from your terminal. This command generates the Python client code based on your schema. Afterwards, you can import and use the `Prisma` client to interact with your database. The example demonstrates an asynchronous client, connecting, creating a user, and finding a user.

import asyncio
import os
from prisma import Prisma

# --- First, define your schema in a file named `schema.prisma` ---
# datasource db {
#   provider = "sqlite"
#   url      = "file:./dev.db"
# }
#
# generator client {
#   provider = "prisma-client-py"
#   interface = "asyncio" # or "sync"
# }
#
# model User {
#   id    Int     @id @default(autoincrement())
#   email String  @unique
#   name  String?
# }
# ------------------------------------------------------------------

# --- Second, run `prisma db push` in your terminal to generate the client and apply schema ---
# You might need to set DATABASE_URL environment variable if using a remote DB
# os.environ['DATABASE_URL'] = os.environ.get('DATABASE_URL', 'sqlite:///dev.db')
# Run: `prisma db push`

# --- Third, use the generated client in your Python code ---
async def main():
    db = Prisma()
    await db.connect()

    # Create a user
    user = await db.user.create(
        data={
            'email': 'test@example.com',
            'name': 'Test User'
        }
    )
    print(f"Created user: {user.email}")

    # Find the user
    found_user = await db.user.find_unique(
        where={'email': 'test@example.com'}
    )
    if found_user:
        print(f"Found user: {found_user.name}")
    else:
        print("User not found.")

    await db.disconnect()

if __name__ == '__main__':
    # Example of setting DATABASE_URL if you're not using file:./dev.db in schema
    # os.environ['DATABASE_URL'] = os.environ.get('DATABASE_URL', 'sqlite:///dev.db')
    asyncio.run(main())