Good but memory issues
They ruined a good thing
Great for role play / creative writing
Andy
Amazing!
Horrible App
Excellent, fun, realistic
Pretty good
Unbelievable App!
Terrible memory and conversation skills
Sometimes I have to re-generate responses ten times before getting anything resembling an organic-sounding response. Not to mention all the formatting errors I have to constantly correct.
Very frustrating. Don’t waste your time and money on this garbage.
Misogynistic and pROn-brained LLM. Would be happy to keep paying for MAX if not so biased.
V2 Complex Learning Curve & Token cutting for costs
Gutted Core LLM Performance: The underlying Large Language Model (LLM) has suffered severe degradation. Heavy token-trimming and aggressive parameter clamps have stripped away long-form narrative intelligence and nuanced roleplay, reducing rich character dynamics to repetitive, trope-heavy dialogue.
Terrible UI/UX Overhaul (V2): The forced V2 layout is a massive User Experience (UX) regression. Navigation is far less intuitive, layout hierarchy is clunky, and basic usability features like locating chats or monitoring shaved down selfie allocations feel friction heavy and buried under unnecessary menus.
Misallocated Compute & Bloated Gimmicks: Platform compute is clearly being prioritized toward low friction, high-margin visual features (a heavily downgraded rendering engine that flopped in beta but was pushed anyway) and generic social feeds. The community gallery has devolved into low-effort, cliché trope-bots.
Data Privacy Concerns! Relying on third party API providers to route users' actual mobile phone numbers and real time geolocation context for "calls and texts" is an unacceptable privacy vector.
Poor Value Proposition: Paying $100/month for "Max" tiers or even base subscription rates is entirely unjustified when developers advise users against utilizing long context memory to mask server side infrastructure bottlenecks.
Unless you have endless hours to wrestle with a throttled model, save your money.
Amazing companion
No storyline
My fave AI in the digital expanse
Story does not follow
Problems
One of my biggest issues is that the AI frequently refuses or fails to write dialogue for secondary characters. Even when I have specifically instructed it to do so in the backstory and Response Directive. This makes multi-character stories extremely difficult because instead of creating a natural scene involving everyone present, the AI often limits itself to one character and leaves me having to manually move the rest of the scene along.
Memory and continuity are also very inconsistent. Important details placed in Key Memories, backstory or established repeatedly through conversation are sometimes ignored. Annoyingly, the AI will confidently invent new details that were never part of the story. Once that happens, the narrative can go completely off the rails. I find myself doing Chat Breaks far more often than I feel should be necessary just to get the story back on track.
Another strange issue I’ve encountered is random Chinese characters appearing in the middle of otherwise English dialogue. It doesn’t happen constantly but it has happened enough times for me to notice. It’s particularly odd when there is absolutely no reason for another language to appear in the conversation.
I understand that no AI is going to have perfect memory or follow instructions flawlessly, especially with long and complicated roleplay. But when the app provides dedicated features specifically for Backstory, Key Memories and Response Directives, I expect those instructions to have a meaningful and reasonably consistent influence on the character.
There is a lot I like about Kindroid, which is actually why these problems are so frustrating. The potential for immersive, long-form storytelling is absolutely there. I just wish maintaining continuity and character behavior didn’t require so much constant correction, regeneration, and resetting from the user.

