Integrations / OpenAI

Trace OpenAI calls

To trace OpenAI calls with Uncoil, wrap your OpenAI client with the Uncoil SDK. Each request is then recorded as a span with the model, token counts, latency, cost and any error.

Read the guideStart free

Your code

OpenAI

Uncoil

One run.

  • 01Model name and parameters
  • 02Input, cached and output tokens
  • 03Latency and errors
  • 04Cost per call, rolled up to the run

Setup

A few lines.

typescript
import OpenAI from "openai";
import { wrapOpenAI } from "uncoil-sdk";

const openai = wrapOpenAI(new OpenAI());

const reply = await openai.responses.create({
  model: process.env.OPENAI_MODEL!,
  input: "Hello",
});

Full details in the documentation. All integrations are listed on the integrations page.

Does tracing OpenAI calls change my responses?
No. Uncoil records each call and sends it to the backend; your request and response are unchanged.

Send your first trace.

Free to start.

Start freeCompare tools