Skip to main content

Multiple lacking features & Unnecessary Rollback on others

Feature RequestWorks/Post/ArticleSearch featureOther

Below are a list of features that I would like requested to be added or changed to match other similar services:
1. Reference images: The “Ai app” version is completely useless in certain art styles and this should already be included in model options as a base feature, similar to nano banana, I should have the freedom to pick the model I want and add reference images especially when ai generations can be inconsistent across varying poses.
Ps. Suggesting to use the completely useless and expensive ai version for character references or nano banana is NOT the point.
2. Profiles are completely useless and there’s been no merit to posting and sharing work unless you already have a large number of followers, this is likely due to censorship and other rollback on features since probably october of 2025.
3. Fix search! The search feature is useless and I cannot find inspiration on seaart or find new authors properly.
4. Heavy censorship on 90% of content on the site, preventing me from even seeing what certain LoRAs even do, why? How am I supposed to select LoRAs if I don’t even know what their results are? Its even applicable on some models where ALL images on them are censored.

Please revert to the older builds which were much easier to use and WERE FUNCTIONAL. Please add the ability to add reference images on all models, other websites provide this feature as a base. I don’t have to use some ridiculous ai app or nano banana or create LoRAs everytime I make a new character generation.

Posted by Sydd ·
Status: Waiting for user response4 comments

Log in to comment and vote

Comments4

  • JJJJulio

    Team•

    Mar 10

    Dear user, thank you for your feedback regarding the issues you mentioned:

    1. We have a corresponding Nano Banana AI APP available for use. Currently, some applications are created and uploaded by users within the platform, and you can look for the ones that suit you best.

    2. We are currently upgrading the personal center. Additionally, you can share your works in the daily challenge and community center to gain more attention.

    3. The search function has been updated and now supports user searches.

    4. We are conducting a secondary review of the relevant content to ensure that works and LoRA that comply with our policies can be seen by more people. Please stay tuned for the latest updates.

    • Sydd

      •

      Mar 10

      Thank you for the response.

      1. Firstly, nano banana and ai apps are not the solution, neither the app nor nano banana are able to reproduce my gens in the same style. I make 4-5 character themed generations per week and your customer support suggested I make LoRA for each? I’m already on the highest tier of subscription, if this feature isn’t even being considered then there really is no merit to using this annual subscription.

      2. Daily challenges rarely, if ever align with what I generate. Back in november, authors had the freedom to make whatever THEY liked and share THEIR works as opposed to reusing prompts for daily challenges from the most popular user.

      4. Everyone wants seaart to be reverted to the pre november/october builds. This solves majority of the dissatisfaction and complains. You’re suggesting further censorship and removal of any and all models that are capable of producing NSFW gens. - Not a solution.

      Please revert to pre november/october 2025 build. Please add reference images as a feature on all models.

      • JJJJulio

        Team•

        Mar 10

        Thank you for your reply. We are currently working hard to restore the feature, but some adjustments have to be made due to external policy reasons.

        Regarding the reference image feature you mentioned, could you please attach a screenshot below?

        • Sydd

          •

          Mar 10

          Yodayo’s Controlnet feature:


          NovelAI’s “Vibe” and “Exact Reference” features:

          Supposedly midjourney, leonardo also provide this feature but I haven’t used either of them.