Inter-2 built for
Social Intelligence
Transcripts capture what was said. Interhuman captures how — voice, face, and body language processed from a single video stream. Build products that act on what's actually happening in a conversation.
Trusted by companies like these
We're building the foundation for socially intelligent AI
Social signal detection — EW-F1
Upper-right is better. Log-scale inference speed.
Customer stories
View all
Built on Behavioural Science
Inter-2 is grounded in behavioral science from the start. Our models are built on research into how people actually communicate, using observable patterns in voice, facial expression, and body language to understand social behavior.
Playthe film
Built for builders
Detect 12 social signals from video, audio, and text, processed together, in temporal alignment. Send a video. Get back detected signals.
Thinking
const form = new FormData();
form.append("file", videoFile);
form.append("model", "inter-2");
form.append("wait_seconds", "60");
const response = await fetch(
"https://api.interhuman.ai/v2/upload/analyze",
{
method: "POST",
headers: { Authorization: `Bearer ${YOUR_API_KEY}` },
body: form,
}
);
const { result } = await response.json();
for (const entry of result.windows) {
console.log(entry.engagement_status, entry.signals);
}