How FanWake works

FanWake vs Fanvue native messages

Fanvue has its own built-in automation — welcome messages for new subscribers, scheduled broadcast messages, and mass DMs. These run entirely inside Fanvue. FanWake does not control or replace them.

FanWake operates separately via Fanvue's official API. It has two distinct engines:

What happens when a fan messages you

From the moment a fan hits send, FanWake runs through a chain of checks before anything reaches your queue or gets sent.

1. 1

   Guard checks

   FanWake first checks: is autopilot enabled for this fan? Has the monthly message limit been hit? Is the review queue full? Is there already a very recent draft for this fan (within 10 minutes)? If any check fails, the message is skipped or the existing draft is superseded.

2. 2

   Debounce wait

   Before generating a reply, FanWake waits a randomised delay to appear human and to batch rapid follow-up messages into a single reply. The wait is chosen at random: 50% chance of 8–30 seconds, 30% chance of 30–60 seconds, 15% chance of 60–120 seconds, 5% chance of 120–180 seconds. Short messages (under 15 characters) add 10 seconds. Long messages (over 80 characters) subtract 10 seconds. After the wait, FanWake checks if this is still the most recent message — if the fan sent something else in the meantime, this draft is cancelled and the newer message starts the debounce again.

3. 3

   Away delay (optional)

   If you have "away delay" enabled, and this is the first message of a cold session (no recent exchange), there is a 35% chance of adding a 10–30 minute delay before the reply appears. This makes the AI feel less like a bot that responds instantly at 3am. It only ever fires on the first message of a cold session — never mid-conversation.

4. 4

   Context assembly

   FanWake reads the last 20 messages from the conversation history, the fan's memory record, your persona settings, the active autopilot strategy, conversation rules, escalation rules, available PPV templates, and which templates this fan has already purchased.

5. 5

   AI generation

   The AI generates a reply in your persona. It determines conversation stage (warming up, building, or engaged), decides whether to pitch a PPV, flags the message for escalation if needed, and optionally drafts a second follow-up message (REPLY_2). It also updates the fan's memory with any new facts the fan stated.

6. 6

   Approval or send

   Based on your approval mode (manual, threshold, or auto), the reply either queues for your review or is sent immediately. If sent, REPLY_2 follows 6–12 seconds later.

Conversation stages

FanWake tracks how deep into a relationship you are with each fan and adjusts the AI's behaviour accordingly. The stage is calculated fresh on every reply.

VIP status is triggered when a fan's total gross on Fanvue reaches €100 (or when manually set). VIP fans are always in the Engaged stage regardless of message count.

Fan memory

Every fan has a persistent memory record that the AI reads before each reply and updates after. It stores facts the fan has directly stated about themselves.

Memory is capped at 2,500 characters. You can view and edit any fan's memory from the Fans page. The AI appends updates at the end of each reply — you don't need to manage it manually.

AI model

The default model is google/gemini-3-flash-preview. It handles the vast majority of conversations well and is fast.

Pro and Elite plans can switch to a different model from Settings. Every model has a built-in fallback: if the primary model declines to generate a reply (for content reasons), a secondary model takes over so the conversation never stalls.

Plan limits

Message limits are hard caps per billing cycle. When the limit is hit, both autopilot and re-engagement pause until the next cycle or until you upgrade. No overages, no surprise charges.

Starter & above plans also unlock: custom conversation rules, custom escalation rules, custom strategy prompts, Telegram notifications, and threshold/auto send. Pro/Elite add AI model choice.