Introducing Humalike Behavioral infrastructure for humanlike AI agents. The social skills & proactiveness your agents have been missing.

Models are capable enough. Smart enough. Fast enough. But agents still feel like AI, so they stall everywhere the experience matters: companions, NPCs, education, voice, etc.

So we built 7 behavioral APIs: → Turn-Taking — when to speak, when to shut up → Theory of Mind — what each person feels → Norms — matches the group's vibe → Persona — opinionated, takes sides → Social Memory — remembers who's who → Social Signals — reads typing, edits, silence → Observability — shows how it landed Model & stack agnostic. Built for groups, not just 1:1.

LoSoNA: a benchmark for social norms in group chats. HUMA: an AI facilitator people couldn't tell from a human.

We are live in PH 👉 We build what makes AI fit in among humans, not just answer them. 💸 $20 in free tokens to start 🔌 One-shot with Hermes, WhatsApp, Telegram & Slack Built by a tiny 🇪🇸×🇵🇱 team that hasn't slept much.

Try now:

Congrats guys!!

🫶🏼

Curious how Norms handles group banter that shifts mid-conversation. Either way, strong launch

congrats teammm 😁👨🏻‍🍳

tysm!

habéis venido a cambiar cosas 👨🏻‍🍳

Let's goooo!! Rooting for you 🚀

Bullish

🫶🏼🫶🏼🫶🏼

crazy

social skills for agents sounds obvious now, but almost nobody is doing it well…

agents don’t need more IQ, they need better social instincts

been waiting for something like this for ages. finally

This is a big step toward AI agents that feel genuinely helpful, not just responsive. Excited to see where Humalike goes!

tysm!

welcome to us Humalike ♥

🫶🏼🫶🏼

La 🐐 de twitter!

best team 🫡🫡

💌

so cool I would use it a lot

This solves a problem most people don't talk about.

That's great

tysm!

Really interesting direction. AI agents need better behavior, not just better intelligence.

Behavioral infrastructure como categoría propia. Creo que esto va a definir cómo hablamos de agentes en 2026

🫶🏼

bangeeeeeer best team w guys

lfg!

💌

Top stuff 🔝📈

@marcusyul good launch team 😁⭐️⭐️

This really touches on a common issue that not many people talk about.

This could be a big step toward making AI agents feel more natural and proactive.

An interesting step toward more natural and proactive AI agents

Humalike team, the 'agents that know when to enter the room' angle is sticky. I made a tiny Saavos front door trained on Humalike so visitors can ask about behavioral APIs, use cases, benchmarks, and stack fit:

vaya pintaza tiene esto

@marcusyul toppp ⭐️⭐️😁

Looks dope!

💌💌

esto es heavyyyy, gran trabajo equipo

GOATS