Venice on Tuesday reduced model session buffers to a maximum of 65000 tokens. There was no notification of this change, yet models are still being advertised as having more memory. This is deceptive. I can understand if regular models need to have a limitation, but pro users should have full access to the models, not a restricted dumbed down version.
If the monthly cost is truly not covering the cost of the needed hardware etc, I personally would be okay with an increase to the $18/month. But paying that, while getting reduced capacity models is not acceptable.
Please authenticate to join the conversation.
New Submission
Feedback
13 days ago

Khaliro
Get notified by email when there are changes.
New Submission
Feedback
13 days ago

Khaliro
Get notified by email when there are changes.