SDK · Python

Minds, idiomatic in Python.

The Python SDK mirrors the TypeScript API and adds first-class MLflow + TensorBoard integration. Sync and async — your choice. pytest examples in the repo. Type-checked with mypy strict.

Start free Read the docs

Capabilities

What you get.

Production-grade from day one. No flags, no betas — these ship when you adopt Minds.

Sync + async

Both `Minds(...)` and `AsyncMinds(...)` — choose your concurrency model.

mypy strict

Full type annotations; works with the latest mypy/pyright.

MLflow + TensorBoard

Reproducibility + experiment tracking integrated.

NumPy-friendly

Embeddings as np.ndarray; zero-copy where possible.

pytest fixtures

First-class test fixtures for in-memory backends.

Server-sent generation

Async generators native — `async for chunk in minds.generate.stream(...)`.

API

A few lines is all it takes.

exampletypescript

from minds import Minds

minds = Minds(api_key=os.environ["MINDS_KEY"])

memories = await minds.recall(
    query="Why did Q3 revenue dip?",
    modes=["vector", "graph", "episodic"],
    timepoint="2026-10-01",
    top_k=5,
)

async for chunk in minds.generate.stream(prompt="Explain the dip."):
    print(chunk, end="", flush=True)

Specs

What it costs, what it guarantees.

At a glance

Package: minds
Python: 3.10+
Types: mypy strict-compatible
ML tracking: MLflow · TensorBoard
Async: asyncio + aiohttp

Guarantees

Durability: fsync per WAL commit
Isolation: MVCC · snapshot
Audit: BLAKE3 Merkle chain
Encryption: AES-256-GCM per-namespace
Concurrency: 100K+ ops/sec

Keep exploring

TypeScript SDK

Same API in TS.

Rust SDK

For zero-overhead native clients.

Continual Learning

What MLflow tracks.

Get started in an afternoon.

Run anywhere — Akasha Cloud, on-prem, or air-gapped.

Start free Talk to sales