Problem feedback

连接 Chatbox AI 失败。这通常是由于配置错误或 Chatbox AI 账户问题。请检查您的设置并验证您的 Chatbox AI 账户状态，或购买Chatbox AI License即可立即解锁所有高级模型，无需任何配置。 API Error: Status Code 400, {"error":{"message":"InvokeModelWithResponseStream: operation error Bedrock Runtime: InvokeModelWithResponseStream, https response error StatusCode: 400, RequestID: 3732ddb0-77e0-41c8-829f-20438bd6c14f, ValidationException: temperature and top_p cannot both be specified for this model. Please use only one. (request id: 202512031427508489115092GLYvdwg) (request id: 202512031427508333160876ploUrx9) (request id: 202512030126595829397198rm4qBwZ)","type":"\u003cnil\u003e","param":"","code":null}}

When using the DeepSeek provider in Chatbox (iOS app v1.18.1 and web version), all requests to DeepSeek fail with a network error, even though the same DeepSeek API key works fine when called via Python. The error is: API Error: Error from DeepSeek: Error: Network Error: Load failed ( api.deepseek.com ) Steps to Reproduce Open the Chatbox iOS app (v1.18.1). Go to Settings → Provider / Model and select DeepSeek as the provider. Enter a valid DeepSeek API key (the same key that works correctly in a Python script calling the official DeepSeek API) Start a new chat using a DeepSeek model (e.g. deepseek-chat). Send any simple message such as “hello”. Observe that the request fails with a network error. The same behavior can be reproduced on the Chatbox web version: Open the Chatbox web app. Go to Settings and configure DeepSeek with the same API key. Start a chat with a DeepSeek model and send a message. Observe the same network error. Expected Results The chat request should be sent to DeepSeek successfully and the model should return a normal response, just like when calling the DeepSeek API via Python with the same API key and base URL. Actual Results • The request in Chatbox (iOS app and web) fails immediately. • Error message shown: API Error: Error from DeepSeek: Error: Network Error: Load failed ( api.deepseek.com ) • No model response is returned. • This happens consistently on different networks and on different devices. Desktop (please complete the following information): • Operating System: N/A (issue is observed on iOS mobile app and web version). • Application Version: • iOS app: Chatbox v1.18.1 • Web app: latest version as of the date of this report Additional Context • iOS version: 17.5.1 • Tested networks: home Wi-Fi and mobile hotspot → the same error occurs on both. • Tested devices: • Original iPhone with Chatbox v1.18.1 • Another phone with a fresh install of Chatbox → same error • I have disabled/cleared any Network Proxy configuration inside Chatbox. • When I open https://api.deepseek.com directly in a browser on the same device/network, it responds (a binary file/response is returned), so DNS and basic connectivity to api.deepseek.com are working. • Using the same DeepSeek API key, a simple Python script calling the official DeepSeek API (e.g. https://api.deepseek.com/v1/chat/completions ) works correctly with no errors. • Therefore, the DeepSeek API itself and my API key appear to be working correctly, and the issue seems specific to the way Chatbox (iOS + web) is integrating with DeepSeek in this environment.

Dear Development Team, I am writing to provide feedback regarding the recent update to Chatbox, version v1.18.1. After updating to this version, I have encountered a potential issue where the application appears to be unable to read or process text content as expected. To diagnose the problem, I reverted to the previous version, v1.17.1. Upon doing so, the text reading functionality returned to normal, and I was able to use the application without any issues. This suggests a possible bug or regression introduced in version v1.18.1 related to text processing. For now, I will remain on version v1.17.1 for stable operation.

I notice there was a charge today that I can't seem to refund, but I'm not sure why since I haven't used the service. Could you please help me understand what might be going on? I'd really appreciate your assistance with this.

{ "error": { "message": "Missing reasoning_content field in the assistant message at message index 2. For more information, please refer to https://api-docs.deepseek.com/guides/thinking_mode#tool-calls ", "type": "invalid_request_error", "param": null, "code": "invalid_request_error" } }

command with image for nano banana 2, reply as folloews: I'm just a language model and can't help with that. word count: 12, token count: 17, tokens used: 20, first token latency: 6882ms, model: Chatbox AI (nano banana 2), time: 21:41

The online search error returned by deepseek-v3.2 might be related to the compatibility issues mentioned in https://api-docs.deepseek.com/guides/thinking_mode#compatibility-notice . I hope the official team can follow up and fix it.

The latest version v1.18.1 can't read the file!

When using some models, the API quota for the current month has already been used up, but at the same time, the API for the next month has been reserved and the date of the next month's purchase will be reset, and it can continue to be used normally. For example, the quota reset time shown above is December 27, 2025, but I am still using it on December 11, and it can still be used. However, the Chat+box+AI quota will show a negative number, and the purchased expansion pack is still 100%, with no signs of being used. What is going on? This is a Chinese translation email, please understand if there are any unclear words