Runtime error when wrapping openAI completion api

[guillaumerybak]

Hey team, we've encountered this error when using the wrapOpenAI + completion api.

dependency versions:

• "braintrust": "2.0.1"
• "openai": "6.10.0"
• "zod": "4.1.13"

completionResponse.withResponse is not a function

TypeError: completionResponse.withResponse is not a function
    at /src/node_modules/.pnpm/braintrust@2.0.1_chokidar@3.6.0_zod@4.1.13/node_modules/braintrust/dist/index.js:9157:72
    at ensureExecuted (/src/node_modules/.pnpm/braintrust@2.0.1_chokidar@3.6.0_zod@4.1.13/node_modules/braintrust/dist/index.js:9171:11)
    at Object.get (/src/node_modules/.pnpm/braintrust@2.0.1_chokidar@3.6.0_zod@4.1.13/node_modules/braintrust/dist/index.js:9182:27)
...
    at /src/node_modules/.pnpm/braintrust@2.0.1_chokidar@3.6.0_zod@4.1.13/node_modules/braintrust/dist/index.js:6607:63
    at AsyncLocalStorage.run (node:internal/async_local_storage/async_hooks:91:14)
    at BraintrustContextManager.runInContext (/src/node_modules/.pnpm/braintrust@2.0.1_chokidar@3.6.0_zod@4.1.13/node_modules/braintrust/dist/index.js:3892:30)
    at withCurrent (/src/node_modules/.pnpm/braintrust@2.0.1_chokidar@3.6.0_zod@4.1.13/node_modules/braintrust/dist/index.js:6607:38)
    at /src/node_modules/.pnpm/braintrust@2.0.1_chokidar@3.6.0_zod@4.1.13/node_modules/braintrust/dist/index.js:5148:18
    at runCatchFinally (/src/node_modules/.pnpm/braintrust@2.0.1_chokidar@3.6.0_zod@4.1.13/node_modules/braintrust/dist/index.js:3142:17)

Do you know already know of this issue or where it might be coming from ?

[cpinn]

Hi,

Could you give me some more information on the code sample you are running?

The completions api returns a promise and after the promise is resolved withResponse is no longer available. It appears you could run into a similar issue if you await the promise and then also await and attempt to access a response.

Correct

const { data, response } = await client.chat.completions
      .create({
        model: "gpt-5-mini",
        messages: [{ role: "user", content: "Say hello" }],
        max_tokens: 10,
      })
      .withResponse()

    const completion = client.chat.completions
      .create({
        model: "gpt-5-mini",
        messages: [{ role: "user", content: "Say hello" }],
        max_tokens: 10,
      })
    const {data, response} = await completion.withResponse();

Incorrect

const completion = await client.chat.completions
      .create({
        model: "gpt-5-mini",
        messages: [{ role: "user", content: "Say hello" }],
        max_tokens: 10,
      });
const  { data, response }  = await completion.withResponse(); // unable to access withResponse no longer available