zespan-fastapi

Automatic LLM observability middleware for FastAPI. Traces every request, LLM call, and custom span with zero code changes for basic usage.

Installation

pip install zespan-fastapi

Quick Start

from fastapi import FastAPI
from zespan_fastapi import ZespanFastAPIMiddleware, ZespanConfig

app = FastAPI()

config = ZespanConfig(api_key="zsp_your_api_key")
app.add_middleware(ZespanFastAPIMiddleware, config=config)

Every HTTP request is now automatically traced.

Trace LLM Calls

from zespan_fastapi import observe_llm

@observe_llm(model="gpt-4o", provider="openai")
async def generate_response(prompt: str):
    response = await openai.chat.completions.create(
        model="gpt-4o",
        messages=[{"role": "user", "content": prompt}]
    )
    return response

Trace Custom Spans

from zespan_fastapi import observe_span

@observe_span("fetch_context", span_type="retrieval")
async def fetch_context(query: str):
    return await vector_db.search(query)

Set Custom Attributes

from zespan_fastapi import set_attribute

@app.post("/chat")
async def chat(request: ChatRequest):
    set_attribute("user.id", request.user_id)
    set_attribute("model", request.model)

Configuration

Get your API key at zespan.com.