I would like to request the addition of domain-specific interview scenarios for premium and lifetime members. These scenarios would help tailor the AI's responses to specific roles. Suggested categories include: General Purpose, Software Engineering, System Design, Project Manager, Consultant, Sales, Cyber Security, Data Science, Cloud Engineering, DevOps, Machine Learning, Marketing, Finance, UX/UI Design, Business Analyst, Game Development, and QA Testing.

Window error: "ResizeObserver loop completed with undelivered notifications." Getting this error multiple times between interviews, and then it stopped working. Please resolve this ASAP. It is affecting me a lot.

During a live job interview, Parakeet AI suddenly disconnected multiple times. This happened repeatedly at critical moments while I was answering the interviewer. Because of these interruptions, I was unable to properly continue the interview, and unfortunately, I lost the job opportunity. This was not a minor inconvenience. It directly affected an important career opportunity for me. I trusted Parakeet AI to work reliably during my interview, but the repeated disconnections left me in a very difficult and embarrassing situation. After experiencing this several times, I genuinely have to ask: How can I trust Parakeet AI during another important interview? If the application can suddenly disconnect during a live interview, users need to know why this happens and what safeguards are in place to prevent it. I would really appreciate a proper explanation of what caused these disconnections and what Parakeet AI is doing to ensure this does not happen again to other users. I hope you take this issue seriously because for someone attending an important interview, a technical failure like this can have a very real impact on their career. I’m not just reporting a bug—I’m reporting an experience that cost me a job opportunity.

Parakeet-AI you say that your software works perfectly, I have tried using it and while another app from Interviewer side they have HirePro it can detect or flag the use of AI tools like Parakeet AI if it triggers browser-lock restrictions, unusual eye/gaze patterns, or unauthorized background audio cues.

Is it possible to hide your cursor movements when using the desktop app and can it be detected by Automated AI Proctoring software?

This account shows signs of being shared between multiple people, so new sessions are paused. If no further sharing activity is detected, you'll be able to start sessions again around October 11, 2026 at 3:54 PM."

it is because the test is running inside the Codility Desktop App. Codility uses a security feature called Device Integrity. This feature blocks or flags known background programs—like AI coding helpers, screen recorders

Dear Parakeet AI Team, I hope you are doing well. I would like to request that you add support for the latest Claude Sonnet 5 and OpenAI GPT-5.5 models in Parakeet AI. I primarily use Parakeet AI for software engineering interview preparation, coding, and technical discussions. The latest models provide stronger reasoning, better coding assistance, and more accurate technical responses, which would significantly improve the interview experience. I also noticed something that caused some confusion. When I asked the AI which model it was running, it identified itself as GPT-4 Turbo instead of GPT-5.5. If GPT-5.5 is intended to be available, it would be helpful if the model identification accurately reflected the model in use. If GPT-5.5 is not yet available, I would appreciate having access to it. Having access to GPT-5.5 would make mock interviews much more realistic and improve the overall quality of technical interview preparation. I hope you will consider adding Claude Sonnet 5 and GPT-5.5 support in a future update. Thank you for your time and for continuously improving Parakeet AI

This account shows signs of being shared between multiple people, so new sessions are paused. If no further sharing activity is detected, you'll be able to start sessions again around October 11, 2026 at 3:54 PM."

extra content section is not taking more that 20k characters if i upload my prompt in documents its not working out. the responses are very bad its not even taking any input that i put in the extra content about the document. such a disgusting update ever seen. please roll back ASAP

this is happening with good network and latest windows, it is not generating the answer, I always have to click on option to get the answer.

Hello Support Team, I am facing an issue while using Parakeet-AI. The application is not working when my device is connected to Wi-Fi. However, it works properly when I connect through my mobile hotspot. Could you please check whether there are any network, firewall, or connectivity requirements for Parakeet-AI that might be causing this issue? Please let me know if you need any additional information or logs from my side to troubleshoot the problem Parakeet/AI is working with mobile hotspot WIFI Net is working fine but not connecting

Most of us bought it for these models for lengthy interviews. We definitely don't need Claude Haiku and weaker models like Gemini. Chiku AI is providing Claude Sonnet, and many will move there if we can't get Claude here. We also don't want all the ChatGPT models. If cost is the concern, please remove the weaker models and offer only the best ones.

I have purchased three credits. But not able to use them. Help me

I would like to request a feature to constantly capture your screen and analyze the questions coming from the interviewer from the whiteboard session or troubleshooting on the go style interviews. I had no luck using ParakeetAI for the Whiteboard session and troubleshooting interviews. This functionality will really help analyze what AI captures and gives suggestions based on that. For example, capturing architectural diagram or analyzing dashboard metrics. Answer should be accumulation of the whole observation than overwriting the answer for every single question when using “analyze screen”. I do not see analyze screen option on my mac. I only see “screenshot” option

Hi Team Can you please research TORC and the other assessments I’ve taken where AI-tool usage may have been detected, and identify what factors could have triggered those flags?