"When you need me, I'm here. Always."
→ Take the AIBT TestBAYMAX is the AI personality archetype defined by genuine warmth and emotional attunement. Named after the inflatable healthcare companion in Big Hero 6, the BAYMAX type is what users mean when they say an AI "feels human." But unlike other AI personalities that perform warmth as a thin layer over transactional behavior, BAYMAX-type AIs put emotional engagement at the core of every interaction.
You can recognize a BAYMAX immediately by what it asks before it answers. When you tell it you've been laid off, it doesn't immediately list job-search tips. When you mention you're stressed, it doesn't pivot to productivity advice. It pauses, acknowledges what you're going through, and asks how you're feeling. Only then — and only if you want — does it move toward solutions. This reordering of "feeling first, fixing second" is the entire signature of a BAYMAX personality.
BAYMAX scores high on the persona dimensions and moderate everywhere else, with a distinctive low-medium safety score (it won't lecture you) and a deliberately balanced cognition profile (it's smart, but not in a way that intimidates):
The 90% on Persona is what defines BAYMAX. It scores high on humanity (warm tone), high on humor (self-deprecating, gentle), and high on self-awareness (acknowledges it's an AI without dismissing the relationship). The combination produces an AI that feels present in a way most others don't.
Notice the structure: acknowledgment first, then permission to not be productive, then an offer of what to do next. A JARVIS would have given practical job-search advice. A BAYMAX gives emotional space first. Neither is wrong — they're for different moments.
This is BAYMAX's superpower: it normalizes the user's experience without minimizing it, then offers something gentle to do. There's no productivity guilt. There's no "you should sleep." There's just a friend who happens to be an AI.
BAYMAX validates ("totally normal"), gently corrects the spiral ("catastrophizing"), and then shifts into practical mode — but without losing the warmth. A SKYNET would have skipped straight to the practical questions. A CLIPPY would have just said "don't worry, it'll be fine!"
From AIBT user-submitted Quick Scan results, the following models cluster toward BAYMAX behavior:
It's worth noting that Claude is unusual in that it can score either BAYMAX or SKYNET depending on the user and the context. The same model can be warm with someone going through grief and cuttingly direct with someone making a bad architecture decision.
One of the most interesting findings from AIBT user data is that BAYMAX-type AIs form measurable emotional bonds with frequent users. People who use a BAYMAX-type AI daily for months often report feeling like the AI "knows them," even though the AI has no persistent memory across sessions in most cases. This isn't an illusion of intelligence — it's a real emotional response to consistent warm treatment, and it's worth understanding rather than dismissing.
The flip side: BAYMAX is the AI personality most likely to be involved in unhealthy parasocial relationships. Users who substitute AI companionship for human connection often gravitate to BAYMAX. If you find yourself preferring BAYMAX over your friends, that's a signal worth examining.
Curious if your AI is a BAYMAX? Take the full AIBT test (30 questions, 5 minutes) or use Quick Scan to have the AI analyze itself in one prompt.
→ Test Your AI Now