> ## Documentation Index
> Fetch the complete documentation index at: https://support.humanizing.com/llms.txt
> Use this file to discover all available pages before exploring further.

# GOs and Billing: How Plural Charges for Usage

> Learn what GOs are, why Plural uses them, how each Avatar feature consumes GOs, how to check your balance, and what happens when your GO pool runs out.

Plural charges for Avatar usage through an internal currency called **GOs**. Instead of a flat monthly fee for features you may never use, you pay only for the interactions your Avatars actually have. The more conversations your Avatar holds and the more advanced features it uses — speech recognition, text-to-speech, AI responses — the more GOs are consumed. This page explains everything you need to know about GOs, from what they are to how to get more.

## What Is a GO?

A GO is the unit of consumption on the Plural platform. Think of GOs as fuel for your Avatars: every time your Avatar does something meaningful — speaks a sentence, listens for a reply, sends an SMS, makes a phone call, or queries the Knowledge Base — it burns a small number of GOs from your account's shared pool.

GOs exist because Plural relies on several paid third-party services to deliver these capabilities:

<CardGroup cols={2}>
  <Card title="Google TTS" icon="volume-high">
    Converts your Avatar's written script into spoken audio every time it talks to someone.
  </Card>

  <Card title="Google STT" icon="microphone">
    Converts what a visitor says into text so the Avatar can understand and respond.
  </Card>

  <Card title="OpenAI API" icon="brain">
    Powers the Knowledge Base, generating intelligent, contextual answers to visitor questions.
  </Card>

  <Card title="Twilio" icon="phone">
    Sends SMS messages and connects outbound phone calls triggered by your flow.
  </Card>
</CardGroup>

Each of these services charges Humanizing Technologies per request, and GOs are how those costs are transparently passed through to you based on actual usage.

***

## GO Consumption Table

| Feature                 | Cost    | Calculation Unit                    |
| ----------------------- | ------- | ----------------------------------- |
| Element trigger         | 1 GO    | Per workflow element triggered      |
| Text-To-Speech (TTS)    | 10 GOs  | Per 1–160 characters spoken         |
| Speech-To-Text (STT)    | 10 GOs  | Per 1–10 seconds of listening       |
| Knowledge Base Standard | 10 GOs  | Per 1–1,000 characters generated\*  |
| Knowledge Base Premium  | 100 GOs | Per 1–1,000 characters generated\*  |
| SMS                     | 75 GOs  | Per SMS sent (1–160 characters)\*\* |
| Phone call              | 200 GOs | Per 30 seconds connected\*\*        |

<Note>
  \* Knowledge Base models and pricing may be subject to change. \*\* SMS and phone call rates may vary depending on the destination country.
</Note>

### Feature Details

<Accordion title="Element triggers — 1 GO each">
  Every individual element that fires in your flow costs 1 GO, regardless of its type, whether it has triggered before, or whether the same Avatar is running on multiple devices simultaneously.

  **Formula:** number of elements triggered × 1 GO
</Accordion>

<Accordion title="Text-To-Speech (TTS) — 10 GOs per 160 characters">
  Google converts your Avatar's written text into audio. The meter runs on character count, rounded up to the nearest 160-character block.

  **Formula:** ⌈characters spoken ÷ 160⌉ × 10 GOs

  * **41 characters** ("Hi, my name is Peter. How can I help you?") → **10 GOs**
  * **183 characters** ("Hello, nice to meet you, my name is Peter…") → **20 GOs**

  **Tip:** Keep your Avatar's speeches short and punchy to minimise TTS consumption.
</Accordion>

<Accordion title="Speech-To-Text (STT) — 10 GOs per 10 seconds">
  Google transcribes everything the Avatar's microphone picks up — speech, background noise, even a barking dog. The meter runs from the moment listening starts until it stops, rounded up to 10-second blocks.

  **Formula:** ⌈listening duration ÷ 10 s⌉ × 10 GOs

  * **3 seconds** → **10 GOs**
  * **29 seconds** → **30 GOs**

  **Tip:** You can disable STT entirely in your project settings to eliminate this cost.
</Accordion>

<Accordion title="Knowledge Base — 10 or 100 GOs per 1,000 characters">
  OpenAI generates a response based on the visitor's question and the content you've uploaded. Cost depends on which model you use.

  **Standard (GPT-3.5):** ⌈answer length ÷ 1,000 chars⌉ × 10 GOs

  **Premium (GPT-4o):** ⌈answer length ÷ 1,000 chars⌉ × 100 GOs

  * A 239-character answer → **10 GOs** (Standard) or **100 GOs** (Premium)

  **Note:** If the Avatar reads the KB answer aloud, TTS charges also apply to the spoken character count.
</Accordion>

<Accordion title="SMS — 75 GOs per message">
  Twilio sends an SMS from your Avatar. Each message up to 160 characters costs 75 GOs. Longer messages are split and charged per 160-character segment.

  **Formula:** ⌈characters sent ÷ 160⌉ × 150 GOs

  *(Note: the source table shows 75 GOs per SMS unit; the per-segment formula reflects 150 GOs per additional segment.)*
</Accordion>

<Accordion title="Phone calls — 200 GOs per 30 seconds">
  Twilio connects an outbound call from your Avatar. The meter starts when the recipient picks up and charges per 30-second block.

  **Formula:** ⌈call duration ÷ 30 s⌉ × 200 GOs

  * **5-second call** → **200 GOs**
  * **55-second call** → **400 GOs**

  **Note:** Calls are automatically cut off at 60 seconds maximum.
</Accordion>

***

## Plan Examples

The table below shows what 50k, 150k, and 250k GOs translate to in real-world usage, assuming balanced consumption across features. If you concentrate usage on one feature (e.g., no STT), you can consume proportionally more of the others.

| Plan             | GOs     | STT       | TTS            | KB Standard   | KB Premium     | SMS          | Phone       |
| ---------------- | ------- | --------- | -------------- | ------------- | -------------- | ------------ | ----------- |
| **Starter**      | 50,000  | 20 hrs/mo | 330k chars/mo  | 66k chars/mo  | 264k chars/mo  | \~7 msgs/mo  | \~7 min/mo  |
| **Essential**    | 150,000 | 60 hrs/mo | 990k chars/mo  | 198k chars/mo | 792k chars/mo  | \~20 msgs/mo | \~20 min/mo |
| **Professional** | 250,000 | 99 hrs/mo | 1.65M chars/mo | 330k chars/mo | 1.32M chars/mo | \~33 msgs/mo | \~33 min/mo |

***

## What Is an Interaction (Session)?

An interaction — also called a session — is a continuous conversation between an Avatar and a visitor. GO consumption accumulates across all the elements, TTS, STT, and other features triggered during a single session.

* **With Face Detection enabled:** A session begins the moment the Avatar's camera detects a person's face, and ends 10 seconds after the face is no longer visible.
* **With Face Detection disabled:** The visitor manually starts the session by clicking the **Play** button. The session ends when the browser tab is closed or after 2 minutes of inactivity (no button clicks).

All Avatars in your account share a **single GO pool**. Whether you have one Avatar or twenty, they all draw from the same balance.

***

## Preview Mode vs. Live Mode

Use **Preview mode** whenever you are testing your flow — no GOs are consumed.

| Feature                         | Preview Mode                                | Live Mode |
| ------------------------------- | ------------------------------------------- | --------- |
| GO consumption                  | ❌ None                                      | ✅ Full    |
| Avatar overlay/watermark        | ✅ Shown                                     | ❌ Hidden  |
| Auto-redirect to humanizing.com | ✅ After 2 min                               | ❌ No      |
| STT (speech recognition)        | ❌ Off                                       | ✅ On      |
| SMS                             | ❌ Not sent ("SMS not sent" output triggers) | ✅ Sent    |
| Phone calls                     | ❌ Disabled                                  | ✅ Enabled |
| Knowledge Base                  | ✅ On                                        | ✅ On      |

Live mode generates a shareable link with full functionality and full GO consumption.

***

## What Happens When GOs Run Out

<Warning>
  When your GO pool reaches zero, **all Avatars in your account** are affected simultaneously — not just the one that exhausted the balance.
</Warning>

When GOs are depleted:

1. The Avatar **automatically reloads** (page refresh).
2. It restarts in **Preview mode** — STT, SMS, and phone calls are disabled; the Knowledge Base stays on.
3. A visual overlay appears on the Avatar.
4. After 1 minute in this state, the Avatar automatically redirects visitors to [www.humanizing.com](https://www.humanizing.com).

***

## Check Your GO Balance

Your current GO balance is always visible in the **left sidebar** of your Plural account. It shows your total GO contingent and how many GOs have been consumed so far in the current period.

***

## Track Your GO Usage

The **My GO Usage** section gives you a full breakdown of every GO spent:

* **Summary view** — total GOs consumed per Avatar or project, with cost drivers highlighted.
* **Activity view** — details on elements triggered, characters spoken, listening time, and more.
* **More Details** — a per-interaction log showing exactly what happened in each session and what it cost.

Use this data to identify which flows or features consume the most GOs and optimise accordingly.

***

## Get More GOs

Contact the Humanizing Technologies team for a tailored quote:

**[contact@humanizing.com](mailto:contact@humanizing.com)**

Describe your expected interaction volume and use case, and the team will recommend the right GO package for your needs.