[roboy]

I really like the increase in transparency, I found it somewhat disturbing to pay for what feels like a random amount of stuff. How should I know if I need Pro or Pro+ if there is no estimate out there what either might get me. The update does not seem to change that though. I would love to have a distribution plotted of how much compute I might expect. Or at least Min/Average/Max run time until disconnect (rn. only Max is known).

[cperry]

I aspire to offer that level of transparency. I am foiled by (1) GPU prices can change randomly on me, and (2) it's hard to convey to a user pricing without giving them a huge incomprehensible price sheet.

All is not lost though, I've got a few irons in the fire that should help resolve those points of feedback over the coming year.

In the meantime, you can always just buy a GCP VM and you have all the certainty you want: https://research.google.com/colaboratory/marketplace.html I find most people don't want that because it's a pain that Colab Pro/Pro+ largely abstracts.

[visarga]

They can benchmark a few architectures (ResNet50, BERT) and tell us how many times we can train a model on a specific level of subscription.