@kurdavar.h We are increasing the monthly transcription limit to 100 hours for Librarians. The increase should be in effect within the next week.
4 Likes
Today is August 4th and I just saw, for the first time, the error message that I have exceeded my monthly transcription limit. This canât be true since I havenât uploaded any audio in August. Unless âmonthlyâ is measured in some way other than a typical calendar month, like âthe last 30 daysâ or something?
We really need tools to track our usage of this feature, and UI alerts to warn us when weâre approaching the limit.
@cjordahl Correct, itâs a 30-day rolling limit, itâs not based on a specific month. We are working to make changes here to allow you to easily track your limit and renewal date and this updates are coming soon.
This is a bit technical, but I hit the same limit so tried to see if I could transcribe it myself with whisperAI. Iâm not very good at Python, but wrote a small script that did it (ran it with VScode). you have to install OpenAI-Whisper and ffmpeg (setting up a path to ffmpeg). The paths are hard coded, but this was my script:
import whisper
model = whisper.load_model(âbaseâ)
audio_file_path = r"c:\temp\chapter6.mp3"
result = model.transcribe(audio_file_path)
print(result[âtextâ])
with open(âchapter6.txtâ, âwâ, encoding=âutf-8â) as file:
file.write(result[âtextâ])
It took maybe 3-4 minutes to run on my notebook. The only thing is, it doesnât split big files into smaller chunks, also I donât think the formatting is quite as nice as what lingq is doing, maybe there are some settings for that or higher quality models do it. Maybe someone from lingq can post a python script that does the transcription so people can do this locally.
1 Like
Nice, that doesnât seem too complicated⌠do you run that phyton script in the browser? Do you think it is possible to automatically transcribe locally and auto upload to LingQ?
Currently, I use the open source Vibe - Transcribe audio / video offline using OpenAI Whisper which seems to be using OpenAI-whisper too. which has these features. It needs to be installed locally.
I ran it in VSCode, which is a free app from MSFT. Then I installed Python extension on it, installed ffmpeg (I had that already), pathed it, and then installed whisper AI (e.g. pip install openai-whisperai). Iâm sure there are other ways to do it.
I would think it could be run from a browers, but the python env is kind of big. In regards to auto uploading it, someone from lingq would have to chime in but I feel like if there was the right http request it could take a mp3 file, break into chunks, then send it to lingq to turn into a lesson. Really just replicating what the server is probably doing, just on the users machine.
If lingq is having trouble with server expense, Iâd suggest making an app or browser plugin that runs whisper-al locally and uploads it. Iâm not a web programmer or Iâd try figuring it out myself.
I did try asking Chatgtp (the new agents) to encode it and boy was that⌠interesting. It sort of worked, but was unreliable.
1 Like
I purchased a one-year premium membership last month. When I upgrade from premium to premium plus, will my membership be upgraded immediately, or do I need to cancel my premium membership first? When I considered upgrading, it seemed like the upgrade would take effect after 2026, but Iâm not sure.
The audio transcript is important to me, so I would like to upgrade to Plus as soon as possible. Will the fee I paid for Premium be applied as a discount when I upgrade to Premium Plus?
@mark
@porthos No need to cancel the Premium plan, you will be switched to a Plus plan right away. Your remaining time from the Premium plan will be automatically refunded.
But I couldnât. It says that I will switch to Plus after one year: â(changing to ⌠Plus on âŚ, 2026)â. I tried cancelling it, but nothing changed. I have to wait a year. I also emailed support@lingq.com about this issue. @zoran
@porthos Thanks. Yes, it was actually my bad what I wrote above. The change takes place on the next renewal date. Since itâs in your case in June next year, I will soon get back to you via the email you sent with a solution. Thanks for your patience.
1 Like
@Zoran - I import a lot of shorts from YT but the video doesnât come in, only the transcript. The folder itself quotes how many minutes and seconds have been imported but clicking on the play button and the video that pops up is empty. Do shorts audio not import? I assume this wonât count towards my limit even though the folder says there is video? Iâm a lifetime member, is this the same audio limit?
As always, many thanks for the help. Jane
@JBelly Yes, when importing shorts, you wonât get a video, only a transcript. Generating lesson audio wonât affect your limit, it only applies to transcription of audio files.
Are you going to allow us to roll over our unused minutes?
Audio Transcription (monthly):
Used 142 of 600 minutes. Renews on September 8.
Also, It would be wise to place the âminutes usedâ more than an inch away from the âDelete Accountâ button.
0.02
1 Like
now you have the answer why they did so. Just to pay more for what we already have , It is so called PREMIUM PLUS
1 Like
the same goes for simplify lessons⌠Not anymore my friends⌠you have to pay more for what you already paid before. THIS IS THE NEW POLICY. I hope they will not block again this message. They already did 3 times in other topics. This is not fair. Not even the chance to express frustration⌠and still like LINGQ
1 Like
Fwiw, ffmpeg 8.0 was released a few weeks ago. They claim to have added whisper net support and it will generate a time synced caption file from an mp3 (or a ton of other formats). I havenât tried to use it yet, maybe someone can document how to use it.
2 Likes
The fact is that all AI functionality comes at a cost. Were you to try to get audio transcription at volume or use 11 labs voices outside of LingQ you would have to pay these services directly. We integrate them into our application for convenience but for those users using a lot of transcription, it seems reasonable to charge an extra fee. As for somehow blocking your comments, that is not something we have ever done.
Glad to hear that Mark. Thanks
1 Like