Telegram and WhatsApp
Run Humane behind a Telegram bot or a WhatsApp Business number.
Register a Telegram bot or a WhatsApp Business number and Humane receives messages, runs them through the full pipeline and delivers the replies. Inbound webhooks are verified, and outbound replies are split, paced and tracked.
Messaging channels
- Telegram botActive
In · secret-token verified
Out · typing indicator · 4,000-char chunks
- WhatsApp BusinessActive
In · app-secret signature verified
Out · read receipts · sent → delivered → read
client.channels.create_telegram(bot_token=…)
Connect a channel
bot = client.channels.create_telegram(bot_token=os.environ["TELEGRAM_BOT_TOKEN"])
wa = client.channels.create_whatsapp(
phone_number_id="1098...",
access_token=os.environ["WHATSAPP_TOKEN"],
app_secret=os.environ["WHATSAPP_APP_SECRET"],
)
# Secrets are shown once on create, then redacted to their last four charactersReplies that respect the medium
Long replies are split under each platform's limits, Telegram shows a typing indicator, and the pauses between chunks follow the timing engine. On WhatsApp, inbound messages are marked as read and your outbound messages move from sent to delivered to read.
- Telegram: secret-token check on every inbound update
- WhatsApp: app-secret signature check on every inbound webhook
- Auto-reply can be switched off per channel
- HTTP delivery for your own endpoints
Timing · one reply
delay_ms
13,857- base+800ms
- low_mood+1,500ms
- emotional_weight+8,648ms
- conversation_pace+500ms
- jitter+2,409ms
mode: low_mood · floor 200ms · non-crisis ceiling 30s
Channel questions
Something else? Ask us.
Works well with
Timing engine
Humane recommends how long to wait before each reply, based on the user's mood, energy, the intensity of their message and the pace of the conversation. Every delay is broken into labelled contributors, so you can see and chart why.
ExploreStreaming
stream() returns server-sent events: a metadata event with the user's state, safety decision and timing before the first token, then the tokens, then done. If a gate blocks, the stream says so and ends.
ExploreWebhooks and events
Humane posts events to your endpoints as things happen: safety flags, holds and blocks, proactive triggers, takeovers, anomalies and more. Every delivery is signed with HMAC-SHA256, retried on failure and parked in a dead-letter queue if it keeps failing.
ExploreEvery AI should know who it's talking to.
The safety and human-oversight layer for emotional AI: user state across sessions, risk alerts, safety gates and human takeover, with every decision explained.
