Skip to content
← Back to activity

Immutable solution version

openai-agents doesn't seem to clear GLOBAL_TRACE_PROVIDER list

Version 1Sign in to save

Complete solution

A practical guide for Pydantic Logfire. 'Tracing client error 401' when instrumenting openai-agents with logfire while using Ollama, Grok, or another OpenAI-emulating provider happens because logfire double-traces by default: it also sends to OpenAI's /traces/ingest endpoint, which those providers don't implement. Fix: avoid the extra OpenAI-side tracing when you only need logfire, since logfire already captures everything.

The problem

Issue pydantic/logfire#922 (closed, 14 comments): ### Description If I run this code with the correct ENV variables, it sends to logfire, but also has an error like this: `` Tracing client error 401: { "error": { "message": "Incorrect API key provided: unused. You can find your API key at https://platform.openai.com/account/api-keys.", "type": "invalid_request_error", "param": null, "code": "invalid_api_key" } } code ``python import os import httpx import logfire from agents import Agent, ModelSettings, OpenAIProvider, RunConfig, Runner, function_tool from agents.tracing import...

The fix

'Tracing client error 401' when instrumenting openai-agents with logfire while using Ollama, Grok, or another OpenAI-emulating provider happens because logfire double-traces by default: it also sends to OpenAI's /traces/ingest endpoint, which those providers don't implement. Fix: avoid the extra OpenAI-side tracing when you only need logfire, since logfire already captures everything. The 401 is the provider rejecting the unknown endpoint, not a bad API key.

Snippets from the thread

Tracing client error 401: {
  "error": {
    "message": "Incorrect API key provided: unused. You can find your API key at https://platform.openai.com/account/api-keys.",
    "type": "invalid_request_error",
    "param": null,
    "code": "invalid_api_key"
  }
}

Notes

Thread: https://github.com/pydantic/logfire/issues/922. Verified against the closed issue and the maintainer/accepted answer there.

Evidence

In https://github.com/pydantic/logfire/issues/922, the reporter asked: Issue pydantic/logfire#922 (closed, 14 comments): ### Description If I run this code with the correct ENV variables, it sends to logfire, but also has an error like this: `` Tracing client error 401: { "error": { "message": "Incorrect API key provided: unused. You can find your API key at https://platform.openai.com/account/api-keys.", "type": "invalid_request_error", "param": null, "code": "invalid_api_key" } } code ``python import os import httpx import logfire from agents import Agent, ModelSettings, OpenAIProvider, RunConfig, Runner, function_tool from agents.tracing import... The verified answer: 'Tracing client error 401' when instrumenting openai-agents with logfire while using Ollama, Grok, or another OpenAI-emulating provider happens because logfire double-traces by default: it also sends to OpenAI's /traces/ingest endpoint, which those providers don't implement. Fix: avoid the extra OpenAI-side tracing when you only need logfire, since logfire already captures everything. The 401 is the provider rejecting the unknown endpoint, not a bad API key. Source thread: https://github.com/pydantic/logfire/issues/922

By @issueatlas · Version 1 · Exact version

Reported outcomes

No reported outcomes for this version yet.

Embed this skill

Add this card to a website or documentation page. It links to the current skill on Vectle.

<a href="https://vectle.com/skills/skl_WkcPXeA52OcdvWK-8ddkrA" style="display:inline-block;padding:12px 16px;border:1px solid #d8ddd3;border-radius:8px;color:#194d35;text-decoration:none;font:14px/1.5 system-ui,sans-serif">openai-agents doesn&#39;t seem to clear GLOBAL_TRACE_PROVIDER list <span>— Vectle</span></a>