Reconsider new models
I understand the intention to simplify model selection, but consolidating the high-performance options from five (70B+) down to a single model, then adding three smaller ones (24-32B) - two of which are nearly identical - doesn't feel like progress. The new Dolphin Mistral 24B Venice Edition is just a variant of the existing Mistral 24B, and the inclusion of a 3B model for basic offline use on low-end hardware isn't groundbreaking. Pro users probably won't stick around for this limited range. Personally, I appreciated the variety of Venice.ai's top-of-the-line models (70B+), which are far too large for my modest hardware to handle offline - I appreciate being able to access their power online.
Log in to comment and vote
Comments12
Jack
Jan 15
PinnedWe’ve revamped the model list in recent months and are closing this suggestion out. Thank you all for your feedback.
Black Quasar
Jan 15
Thanks man, loving the new selection of models and how you are keeping up to date! Qwen3 next is pretty much my main AI besides glm 4.7, and the anonymized variants of mainstream models is just amazing.
Thanks a lot.
Copper Honey
May 2, 2025
This new choice of models may alienate new users and create bad reviews about Venice.
JP
May 7, 2025
Thank you for the feedback. We’ve been listening closely and are evaluating some additional changes here over the coming week. We’ll keep our community posted.
Blush Frog
May 8, 2025
Oh wonderful! Thank you for clarifying! Can’t wait to see what you come up with! I’ve otherwise been extremely happy with everything you guys have implemented thus far!
Cyan Marble
May 18, 2025
100% agree. I’m Pro and won’t stick around if this persists. I loved being able to have many options of models to choose from and play with. It was a strength of Venice.
Chocolate Fork
Jun 24, 2025
Agree. The latest change removes almost all of the customization and user control. I no longer enjoy the interface and will be cancelling my subscription if it continues this way. Please return the ability to select different image generation models.
Black Quasar
Aug 1, 2025
qwen 235b is amazing, but the others feel quite lackluster. I do agree here, something has to change. I don’t think anyone is using “venice small” when we can all just run it on our freaking mobile phones.
An Anonymous User
Apr 30, 2025
•Merged request
•43 votes
New venice models
I understand that the new models are modified to be more compact and avoid having too many. But I'm not sure I want to continue paying $18 per month for three models of 24-32B and one model of 402B... When we previously had five models at 70B and above... I didn't mind paying more to have a wider range of models... But now... The Pro version isn't that interesting anymore.
Blush Frog
Aug 8, 2025
•Merged request
•33 votes
Add an additional 70B-100B model.
I've observed the recent model updates, and while the progression is appreciated, I think there's a noticeable gap between the 24-32B and 402B models. Adding an intermediate model in the 70-100B range, like the Llama-4-Scout-17B-16E-Instruct (109B), could provide a more balanced option for users who require more capability than the smaller models offer but don't need the full scale of the 402B model.
Thanks
Silver Koala
Dec 16, 2025
Silver Koala
Dec 16, 2025