Hi @david ,
It is working with Chrome on my MacAir.
When I tried with my iPhone8 it did not (even with Chrome Browser). Have a clue?
All the best,
Edu
On iOS, all browsers are actually Safari under the hood.
2 Likes
Absolutely frigging amazing!!! Excuse the casual nature of my comment but am really impressed and excited about using this feature and the new possibilities it brings
1 Like
Not yet - fully testing all other cases next week. What do you think is the reason for the performance gap?
Thanks!
I don’t know whether there is a performance gap.
We’re thrilled with the Whisper integration and believe it has great potential for building new tools. However, we’ve run into an issue. Despite recording audio of various lengths, only one second of it is being utilized. When we check the ‘Usage’ tab on the Open AI website, all Whisper requests are also logged as one-second long.
We’re unsure if this is due to a limitation in audio processing on your end or if it’s a bug with Open AI’s Whisper. Could you please help us identify and resolve this problem?
Hi, did you generate the speech to text within the data column or on action trigger? Also will you be able to provide a snapshot of your data ?
1 Like
What if you use it as an action instead of a column?
The same
I can’t reproduce it. I uploaded a 3:22 video of a person speaking words, and it does work for longer than 1 second.
I ran into a similar issue and realized that files are translated but not inputs from the glide voice recorder. There is a difference in the file format m4a & mp4 but both work according to OpenAI. However, if you open them from the data table, the glide recorder version is opened as a link and then downloaded so it might be a storage issue. Found this in the OpenAI FAQ which might be the answer:
Can I send links to audio files instead?
We don’t support this right now. You’ll need to send a file in one of the formats listed above.
1 Like
Have you tried not to upload the file, but to record it directly in the application? As on the example of David
@ThinhDinh Hi, I am not sure what I am doing wrong, but the speech to text is not working for me. I kindly ask that you help me look at the table I created below. Is there something I am supposed to do.
Thanks for your help
Any help please. I am still stuck. Is there anyone that can assist?
What file extension are you seeing for your audio file?
Many thanks @ThinhDinh.
It’s MP3. I used Glide recorder from Chrome browser.
Something like this:
Thank you.
I assume you have already connected your API key with the OpenAI integration, and you have a card tied to OpenAI?
Yes. I copied the API key from ChatGPT and pasted in the integration, but I am not sure what ‘a card tied to OpenAI’ mean.
How can I go about doing that?