Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.usetitan.app/llms.txt

Use this file to discover all available pages before exploring further.

WhatsApp’s anti-spam systems are designed to detect automated behavior. Accounts that send messages too quickly, without any typing state, or without ever going online and offline, can attract restrictions or bans. Safe Mode tells Titan to behave more like a real person: it types before sending, waits a realistic amount of time, and goes online before a message and offline after.
Safe Mode adds intentional delays to every send operation. It reduces throughput significantly. Do not enable it when you need maximum sending speed.

​
What Safe Mode does

When Safe Mode is enabled for a session, Titan automatically applies the following behaviors before and after each send:
Before sending a text message, Titan automatically sends a composing typing state to the recipient. This makes the interaction look like a person is actually typing.
The delay before sending is calculated from the character count of the message at approximately 3.3 characters per second, with a minimum of 800ms and a maximum of 12 seconds. A short message gets a short delay; a longer one gets more time.
Before sending an image, video, document, or sticker, Titan waits 2 seconds with random jitter to simulate the time it takes a real user to select and upload a file.
Before sending an audio voice note, Titan sends a recording indicator and waits 3 seconds, simulating the time a person would spend recording.
Titan sets the session’s presence to online before sending the message and offline again afterward. This mirrors the pattern of a real user picking up their phone, sending a message, and putting it down.
All delays have a random multiplier between 0.7× and 1.3× applied. This prevents perfectly uniform timing that could be fingerprinted as automation.

​
Enabling Safe Mode

Safe Mode is configured per session. Pass safeMode: true when creating a session or when updating an existing one: 
POST /api/sessions
Authorization: Bearer titan_...
Content-Type: application/json

{
  "name": "support",
  "safeMode": true
}

​
Proxy configuration

Safe Mode pairs well with per-session proxy configuration. Routing a session’s WhatsApp traffic through a consistent IP address makes the account look like it belongs to a real device in a fixed location, rather than a server with a rotating IP. You can configure a proxy when creating or updating a session: 
PUT /api/sessions/{session}
Authorization: Bearer titan_...
Content-Type: application/json

{
  "proxy": {
    "url": "socks5://user:pass@host:port"
  }
}
Titan supports SOCKS5 and HTTP proxies. For managed proxy pools, Smartproxy and Oxylabs are supported providers with built-in integration.
For new phone numbers or accounts you’re warming up, enable both Safe Mode and a dedicated proxy from day one. Establishing consistent behavioral patterns early reduces the risk of restrictions later.

​
Throughput impact

The table below shows approximate throughput when sending text messages of different lengths. All numbers assume the maximum jitter multiplier (1.3×).
Message lengthDelay rangeMax sends per minute
10 characters800ms – 1.04s~45–50
50 characters~1.4s – 1.9s~25–30
150 characters~3.4s – 4.5s~10–12
400+ charactersUp to 12s~4–5
These are single-session numbers. You can run many sessions in parallel — each session’s Safe Mode delays are independent.

​
When to use Safe Mode

  • New numbers: always enable Safe Mode when warming up a newly linked number.
  • Production support sessions: where messages are sent in response to real users, Safe Mode makes responses look authentic.
  • Long-term accounts: accounts that need to stay active for months or years benefit from consistently human-like patterns.
Safe Mode is not necessary for purely internal or low-volume sessions where account longevity is not a concern, or when you need maximum throughput and accept the associated risk.