mixpeek 1.3.6

Mixpeek API

Mixpeek Python SDK

The official Python client for the Mixpeek API.

Install

pip install mixpeek

Quickstart

from mixpeek import Mixpeek

client = Mixpeek(api_key="YOUR_API_KEY")

# Create a standalone namespace
client.namespaces.create(namespace_id="my-search", mode="standalone")

# Upsert documents with your own vectors
client.namespaces.documents.upsert(
    namespace_id="my-search",
    documents=[{
        "document_id": "doc-001",
        "vectors": {"embedding": [0.12, -0.34, 0.56]},
        "payload": {"title": "First document", "category": "demo"},
    }],
)

# Search (vectors are searchable within ~10-30 seconds of upsert)
results = client.search(
    namespace_id="my-search",
    queries=[{
        "vector_name": "embedding",
        "vector": [0.15, -0.28, 0.44],
        "top_k": 10,
    }],
)

Authentication

Pass your API key directly or set the MIXPEEK_API_KEY environment variable:

# Explicit
client = Mixpeek(api_key="sk_...")

# From environment
import os
os.environ["MIXPEEK_API_KEY"] = "sk_..."
client = Mixpeek()

Namespace-scoped operations

Set a default namespace to avoid passing it on every call:

client = Mixpeek(api_key="sk_...", namespace="ns_abc123")

# Uses the default namespace
results = client.search(query="red shoes", collection="products", limit=20)

Resources

The client exposes resource managers for the full API:

client.namespaces            # create, list, get, delete
client.namespaces.documents  # upsert (BYO vectors)
client.buckets               # create, list, get, delete, upload
client.collections           # create, list, get, delete, trigger
client.retrievers            # create, list, get, delete, execute
client.documents             # list, get, delete, update, search

Requirements

• Python 3.9+
• No heavy dependencies (uses urllib3 for HTTP)

Links

• Documentation
• API Reference
• Quickstart Guide
• GitHub