Skip to content
Open
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion packages/proxy/src/proxy.ts
Original file line number Diff line number Diff line change
Expand Up @@ -1139,7 +1139,7 @@ async function fetchModelLoop(
headersString.push(`${key}: ${value}`);
});
const errorText =
`AI provider returned ${httpCode} error.\n\nHeaders:\n` +
`The AI provider returned ${httpCode} error. This is likely not an issue with Braintrust, but rather with the model provider.${httpCode === 500 ? " Please check the model provider's status page." : httpCode === 429 ? " Please adjust max concurrency." : ""}\n\nHeaders:\n` +
Copy link
Collaborator

@ibolmo ibolmo Jun 10, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The "please adjust concurrency" is throwing me off. Is that on the Braintrust side (i.e. playground concurrency) or is that like a quota/limit problem in the AI provider?

You may consider just dropping this one and keep the rest. I think most folks know what 429 is.

Would you see us return the statusText and/or response.text() in addition to headers?

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm also curious if we shouldn't return a JSON message rather than a plain message. This way we avoid losing fidelity.

headersString.join("\n");
proxyResponse = {
response: new Response(null, { status: httpCode }),
Expand Down
Loading