Skip to content

Shakti vs FastAPI

FastAPI and Shakti are both async, ASGI-based Python web frameworks, but they solve different problems by design. This is a structural comparison, not a benchmark — we haven't published performance numbers for either, and you shouldn't trust a framework comparison that invents them.

Where FastAPI is deliberately minimal

FastAPI's core strength is request/response validation and automatic OpenAPI documentation, built on Pydantic. It doesn't ship an ORM, an auth system, or an admin panel — and that's intentional. You choose SQLAlchemy or Tortoise or nothing at all; you choose your own auth library; you assemble the stack that fits your project.

That's a real advantage when you already have strong opinions about your ORM, your auth flow, or your deployment shape, and don't want a framework making those choices for you.

Where Shakti is opinionated

Shakti bundles the pieces most APIs end up needing anyway:

from shakti import Shakti, Depends
from shakti.orm import Database
from shakti.auth import Auth
from shakti.admin import Admin

app = Shakti(title="My App")
db = Database("sqlite+aiosqlite:///./app.db")
db.init_app(app)

auth = Auth(db, secret_key="...")
auth.init_app(app)

admin = Admin(db, auth)
admin.init_app(app)

That's a database connection, JWT authentication, and a working admin UI, with no separate library choices to make. See Getting Started.

One honest gap: request validation

FastAPI's Pydantic-based request validation is genuinely strong, and it's worth naming directly: Shakti doesn't have an equivalent typed-schema validation layer yet. Handlers take a plain dict body today — see Request & Response. If automatic request-schema validation with field-level error messages is the feature you rely on most, that's currently a point in FastAPI's favor.

Where Shakti goes further

  • Admin panel — register a model, get a working CRUD UI. FastAPI has no built-in equivalent.
  • AI integration — chat, streaming, RAG, and tool-calling agents are first-class, not a third-party add-on. See AI Overview.
  • Background jobs — an async job queue with retries and scheduling is built in. See Workflows.
  • Auth + RBAC — JWT, refresh tokens, and role-based access control ship ready to use. See JWT Auth.

Which one fits

If you want maximum control over every dependency and already have a validation/ORM/auth stack you like, FastAPI's minimalism is a feature. If you want an API, an admin UI, and AI features running from a single install without assembling them yourself, Shakti is built for that. Both are legitimate choices — they're just optimizing for different defaults.