[Bug] Models dont work anymore. Only Auto works

[J_M]

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

Since the last few versions, choosing a model (other than Auto) doesnt work anymore. The request starts, then it stops suddenly. Sometimes with no message/error at all, other times saying that I need to buy credit. Credit for what exactly? Where is my usage count? nothing in my account says Im over a limit, and why is that true for any model I choose? what am I paying $20 for then exactly?

I tried to add my own key for e.g. anthropic, same thing, selecting any claude model still does nothing, doesnt work.

What are we paying for exactly? a pretty ide? we should be able to see overage usage when/if actually exceeded and own api should always override that.

Nothing in my billing profiles says Ive gone over whatever calls or credit is included, so it’s impossible to know when or how I went over, if that’s even true at all.

It’s starting to get ridiculous… every release degrades the core features of Cursor, rather than improving. Nobody cares about new features if all the other ones stop working.

Steps to Reproduce

Choose any model besides Auto. Even with own api key, nothing works, only Auto.
Sometimes it prompts to spend more money without showing why or how credit has been exceeded/when.

Expected Behavior

Choosing a model would work just fine before…

Operating System

Windows 10/11

Current Cursor Version (Menu → About Cursor → Copy)

Version: 2.2.43 (user setup)
VSCode Version: 1.105.1
Commit: 32cfbe848b35d9eb320980195985450f244b3030
Date: 2025-12-19T06:06:44.644Z
Electron: 37.7.0
Chromium: 138.0.7204.251
Node.js: 22.20.0
V8: 13.8.258.32-electron.0
OS: Windows_NT x64 10.0.26200

Does this stop you from using Cursor

Sometimes - I can sometimes use Cursor

[deanrie]

Hey, thanks for the detailed report.

Based on your description, this might be related to hitting your monthly usage limit, but the UI isn’t showing it correctly. Let’s check a few things:

1. Check usage on the dashboard: Go to https://cursor.com/dashboard?tab=usage and take a screenshot. The Pro plan includes $20 of API usage, so it’s important to confirm whether you’ve hit the limit.

2. Check the API setting: Settings > Models > API Keys. Make sure “Override OpenAI Base URL” is turned off. This setting can cause issues with Anthropic keys even if the key is valid.

3. Request ID: Next time the error happens, copy the Request ID (chat context menu > Copy Request ID). This will help the team investigate.

4. Exact error: Please take a screenshot of the error message when it appears.

Share that info and we can narrow down the issue.

[J_M]

As mentioned, nothing in my Usage tab says anything about how much Ive used, how much to maxing out or if it’s already been maxed out. It’s pointless to send a screenshot, it doesnt say anything, all I see is a list of requests and it all says included.

Also, using a model with my own api key doesnt work, nothing happens (it shows in the requests, but as 0 tokens).

{“error”:{“type”:“provider”,“reason”:“provider_error”,“message”:“Provider returned 401”,“retryable”:false,“provider”:{“status”:401,“body”:“{\n “error”: {\n “message”: “Incorrect API key provided: sk-ant-a************************************************************************************************gQAA. You can find your API key at https://platform.openai.com/account/api-keys.\”,\n “type”: “invalid_request_error”,\n “code”: “invalid_api_key”,\n “param”: null\n },\n “status”: 401\n}”}}}

Request ID: 7c0fde13-b9a4-4d81-8422-c0ed330cd266
{“error”:“ERROR_OPENAI”,“details”:{“title”:“Unable to reach the model provider”,“detail”:“We encountered an issue when using your API key: Streaming error\n\nAPI Error:\n\n\n{\"error\":{\"type\":\"provider\",\"reason\":\"provider_error\",\"message\":\"Provider returned 401\",\"retryable\":false,\"provider\":{\"status\":401,\"body\":\"{\\n \\\"error\\\": {\\n \\\"message\\\": \\\"Incorrect API key provided: sk-ant-a************************************************************************************************gQAA. You can find your API key at ``https://platform.openai.com/account/api-keys.\\\``",\\n \\\"type\\\": \\\"invalid_request_error\\\",\\n \\\"code\\\": \\\"invalid_api_key\\\",\\n \\\"param\\\": null\\n },\\n \\\"status\\\": 401\\n}\"}}}\n”,“additionalInfo”:{},“buttons”:,“planChoices”:},“isExpected”:true}

Im using claude, not openai, so im not sure why the openaai key even matters here, when it’s actually disabled in my settings.

I swapped the openapi key with a new one, just cause…. i dont even get that popup anymore, chat requests just starts then stops immediately.

It seems there’s a massive bug in models, where even if you turn off openai api key completely, and use some other provider, it still checks for openapi key….

[deanrie]

Еhanks for the details. I can see the API Keys settings screenshot and the “Provider returned 401” error. A 401 here usually means your Anthropic key is being sent to OpenAI because “Override OpenAI Base URL” is enabled.

What to try next:

1. Settings > Models > OpenAI: turn off “Override OpenAI Base URL” and completely clear the URL field
2. Fully quit Cursor, then open it again
3. Temporarily disable your Anthropic key and test Claude via the included Pro access, then re-enable your key
4. If the toggle turns itself back on after restart, or Claude still doesn’t respond, that’s a known issue and the team is working on a fix

Could you please confirm:

• Does “Override OpenAI Base URL” turn itself back on after restart?
• New Request IDs for the error with your key and for the test using the included Pro access

I’ll follow up once I have those details.