Python SDK

What is it?

A thin Python client for BotConnector Core — the same service the desktop app and CLI use. It is not a second inference engine: model loading, runtimes and serving stay inside Core; the SDK talks to its local API.

Source usage (preview)

pip install botconnector-ai

The package is source-only for now; do not use pip install botconnector-ai because no PyPI release exists. The SDK calls the local API and does not add an inference engine.

Target design

from botconnector import Client

client = Client()

models = client.models.list()
model = client.llm("qwen/model")
response = model.chat("Hello")
print(response.text)

In this section

• Python chat
• Python streaming
• Python model management
• Python embeddings
• Python tools

Use today

The local API is OpenAI-compatible, so the standard OpenAI client works:

from openai import OpenAI

client = OpenAI(base_url="http://127.0.0.1:11435/v1", api_key="not-needed")
resp = client.chat.completions.create(
    model="<publisher>/<model>",
    messages=[{"role": "user", "content": "Hello"}],
)
print(resp.choices[0].message.content)

See OpenAI-compatible API for the endpoint reference.

Status

Tracked as Beta / Implemented / Preview on the status page. PyPI publication is not available yet.