When I used the customed api (DashScope of Alibaba Cloud), I tried to employ the third-party api service from the Moonshot of the kimi/kimi-k2.5. However, the cursor raised an error after outputting a few words and calling reading tools a few times, as follows:
Provider returned error: {“error”:{“message”:“thinking is enabled but reasoning_content is missing in assistant tool call message at index 3”,“type”:“InvalidParameter”,“param”:null,“code”:“InvalidParameter”}}
I also tried other third-party models, such as MiniMax/MiniMax-2.5 and MiniMax/MiniMax-2.7 of the DashScope api, these models can run correctly.
Steps to Reproduce
Deploy customed api in the cursor editor.
Add kimi/kimi-k2.5 model.
Select kimi/kimi-k2.5 and talk with the model in the chat.
The error is coming from the DashScope API. When thinking mode is enabled for the kimi-k2.5 model, it requires the reasoning_content field in assistant messages that include tool calls. Right now, Cursor doesn’t preserve that field when it sends the conversation history back to the provider during multi-turn tool-call flows, so the request gets rejected.
This is a known limitation for thinking or reasoning models when using custom API keys. I’ve shared this with the team, but there’s no ETA for a fix yet.
As a temporary workaround:
Use models without thinking mode through DashScope (for example, the MiniMax models you mentioned work fine)
Or check if DashScope has an option to disable thinking mode for kimi-k2.5, which may fully avoid the issue
When tried to use Kimi 2.6 API as a custom model inside Cursor IDE , after the first tool call, i got this error:
Provider returned error: {“error”:{“message”:“thinking is enabled but reasoning_content is missing in assistant tool call message at index 3”,“type”:“invalid_request_error”}}
Steps to Reproduce
Add Kimi api key and “kimi-2.5” as a custom model and try to send a message requiring tool calls
For AI issues: add Request ID with privacy disabled
Request ID: f6b2f1e2-6541-4958-9651-98650cb128f1
{“error”:“ERROR_PROVIDER_ERROR”,“details”:{“title”:“Provider Error”,“detail”:“Provider returned error: {"error":{"message":"thinking is enabled but reasoning_content is missing in assistant tool call message at index 3","type":"invalid_request_error"}}”,“isRetryable”:false,“additionalInfo”:{},“buttons”:,“planChoices”:},“isExpected”:true}
Provider Error Provider returned error: {“error”:{“message”:“thinking is enabled but reasoning_content is missing in assistant tool call message at index 3”,“type”:“invalid_request_error”}}
Svs: Provider Error Provider returned error: {“error”:{“message”:“thinking is enabled but reasoning_content is missing in assistant tool call message at index 3”,“type”:“invalid_request_error”}}
at mg_ (vscode-file://vscode-app/usr/share/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:29432:24260)
at dg_ (vscode-file://vscode-app/usr/share/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:29432:23144)
at wg_ (vscode-file://vscode-app/usr/share/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:29433:6490)
at nCh.run (vscode-file://vscode-app/usr/share/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:29433:11307)
at async lQi.runAgentLoop (vscode-file://vscode-app/usr/share/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:41749:12418)
at async mim.streamFromAgentBackend (vscode-file://vscode-app/usr/share/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:41817:12344)
at async mim.getAgentStreamResponse (vscode-file://vscode-app/usr/share/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:41817:19567)
at async hCt.submitChatMaybeAbortCurrent (vscode-file://vscode-app/usr/share/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:29548:16906)
at async Object.Xa [as onSubmit] (vscode-file://vscode-app/usr/share/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:40804:4244)
at async vscode-file://vscode-app/usr/share/cursor/resources/app/out/vs/workbench/workbench.desktop.main.js:40803:102189
Hey, the issue is still the same. Cursor isn’t passing reasoning_content back through the generic OpenAI-compatible BYOK path, and providers like DashScope, Moonshot, and the direct Kimi API, where thinking mode requires this field in assistant tool-call messages, reject multi-turn requests with tool calls.
We’re tracking it on our side, but I can’t share an ETA for a fix yet.
Current workaround:
Turn off thinking mode for kimi-k2.5/2.6 on the provider side, if your custom endpoint settings allow it.
Use non-thinking models via BYOK, for example MiniMax via DashScope works fine.
If you can, use Kimi through Cursor’s first-party routing instead of your own API key, since reasoning_content is forwarded correctly there.
If there’s an update on the fix, I’ll reply in the thread.
When I use my own API key, I get the following error. (Kimi-2.6)
Request ID: f8e77e31-d5ef-49af-8b1c-e46a6d6359f3
{“error”:“ERROR_PROVIDER_ERROR”,“details”:{“title”:“Provider Error”,“detail”:“Provider returned error: {"error":{"message":"thinking is enabled but reasoning_content is missing in assistant tool call message at index 3","type":"invalid_request_error"}}”,“isRetryable”:false,“additionalInfo”:{},“buttons”:,“planChoices”:},“isExpected”:true}
Provider Error Provider returned error: {“error”:{“message”:“thinking is enabled but reasoning_content is missing in assistant tool call message at index 3”,“type”:“invalid_request_error”}}
des: Provider Error Provider returned error: {“error”:{“message”:“thinking is enabled but reasoning_content is missing in assistant tool call message at index 3”,“type”:“invalid_request_error”}}
at g8y (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:28853:25419)
at p8y (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:28853:24303)
at C8y (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:28854:6490)
at e4d.run (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:28854:11311)
at async Ups.runAgentLoop (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:41180:14333)
at async C0h.streamFromAgentBackend (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:41248:12259)
at async C0h.getAgentStreamResponse (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:41248:19484)
at async Mbt.submitChatMaybeAbortCurrent (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:28969:17002)
at async za (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:40232:4234)
Steps to Reproduce
When I use my own API key, it works in agent mode as long as the K2.6 model is selected.
Request ID: 67ca6aac-36e4-415f-9946-e4da5762c2ec
{“error”:“ERROR_PROVIDER_ERROR”,“details”:{“title”:“Provider Error”,“detail”:“We’re having trouble finding the resource you requested. If the problem persists, please contact support.”,“isRetryable”:false,“additionalInfo”:{},“buttons”:,“planChoices”:},“isExpected”:true}
Provider Error We’re having trouble finding the resource you requested. If the problem persists, please contact support.
VMr: Provider Error We’re having trouble finding the resource you requested. If the problem persists, please contact support.
at Umk (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:35795:26696)
at M2w (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:35795:25580)
at Wmk (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:35796:6852)
at Pdf.run (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:35796:11731)
at async xZs.runAgentLoop (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:48162:21860)
at async YVf.streamFromAgentBackend (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:48230:12345)
at async YVf.getAgentStreamResponse (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:48230:19576)
at async Uyn.submitChatMaybeAbortCurrent (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:35928:17397)
at async Th (vscode-file://vscode-app/Applications/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js:47214:3889)
Hey, @furki, the new Request ID now shows a different error: “We’re having trouble finding the resource you requested”. This isn’t the same reasoning_content issue, it’s a separate bug in how the model slug is parsed for kimi-k2p6 via BYOK. It showed up in a recent agent build. We’ve reported it on our side.
On the original reasoning_content issue, the status is unchanged. It’s being tracked, but I can’t share an ETA for the fix yet. The workarounds from the post above are still valid:
Turn off thinking mode on the provider side, if the settings allow it
Use non-thinking models via BYOK
If possible, use Kimi through Cursor first-party routing, since reasoning_content is passed through correctly there
If you want, send a screenshot of your custom model settings, specifically how the slug for k2.6 is configured. I can check if there’s a local workaround for the “resource not found” error until it’s fixed on the backend.