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

# Pricing Tiers

> Four pricing tiers with usage limits and feature access control

## Overview

Genie Helper offers four pricing tiers designed to scale from individual creators to professional studios. Each tier has specific usage limits based on server resource costs and operational overhead.

## Pricing Structure

| Tier        | Monthly Cost | Target User         | Key Feature                       |
| ----------- | ------------ | ------------------- | --------------------------------- |
| **Starter** | \$0          | Trial users         | Basic platform access, no video   |
| **Creator** | \$49         | Individual creators | Unlimited scheduling, 3 platforms |
| **Pro**     | \$149        | Power users         | Unlimited publishing, 6 platforms |
| **Studio**  | \$499        | Teams & agencies    | 5 team seats, all 9 platforms     |

## Cost Basis

* **Server**: \$100/mo fixed (IONOS dedicated, self-hosted)
* **LLM inference**: CPU-bound Qwen2.5 7B \~2-5s/call, \~4.8GB RAM pinned
* **Stagehand sessions**: \~300MB RAM/active browser session → \~33 concurrent sessions ceiling
* **FFmpeg clip generation**: \~30s CPU per clip (primary bottleneck)
* **Watermark operations**: \~100ms (Sharp/ImageMagick, effectively zero cost)
* **Platform universe**: 9 total platforms supported

## Tier Limits Comparison

<Note>
  `-1` indicates unlimited usage
</Note>

### Starter (Free)

Ideal for testing the platform with minimal resource commitment.

| Feature                  | Limit     | Reasoning                                  |
| ------------------------ | --------- | ------------------------------------------ |
| **AI-Assisted Posts**    | 3         | Taste of LLM capability, minimal CPU load  |
| **Posts Published**      | 5         | Not a production workflow, just testing    |
| **Scheduled Queue Size** | 3         | Friction mechanic — forces frequent logins |
| **Platforms Connected**  | 1         | Single platform test connection            |
| **Taxonomy AI Calls**    | 5         | Teaser of proprietary classification       |
| **Watermark Operations** | Unlimited | Zero cost — drug dealer hook               |
| **AI Clip Operations**   | 0         | Video blocked (CSAM risk + bandwidth)      |
| **Thumbnail Operations** | 0         | Requires video frame extraction            |
| **Team Seats**           | 1         | Solo only                                  |

### Creator (\$49/mo)

Built for individual creators managing multiple platforms daily.

| Feature                  | Limit     | Reasoning                                  |
| ------------------------ | --------- | ------------------------------------------ |
| **AI-Assisted Posts**    | 30        | \~1 per day, acceptable CPU load           |
| **Posts Published**      | 90        | 3 platforms × \~1/day = 90                 |
| **Scheduled Queue Size** | Unlimited | **Primary upgrade hook** from Starter      |
| **Platforms Connected**  | 3         | Most creators focus on 2-3 platforms       |
| **Taxonomy AI Calls**    | 75        | \~2.5 per day for content classification   |
| **Watermark Operations** | Unlimited | Zero marginal cost                         |
| **AI Clip Operations**   | 10        | Aligns with 5GB storage (\~500MB avg clip) |
| **Thumbnail Operations** | Unlimited | Cheap FFmpeg frame extract                 |
| **Team Seats**           | 1         | Solo creator                               |

### Pro (\$149/mo)

Power users with high publishing volume across multiple platforms.

| Feature                  | Limit     | Reasoning                                   |
| ------------------------ | --------- | ------------------------------------------- |
| **AI-Assisted Posts**    | 150       | Heavy user load, self-hosted can handle     |
| **Posts Published**      | Unlimited | Zero marginal Stagehand cost once connected |
| **Scheduled Queue Size** | Unlimited | No artificial friction                      |
| **Platforms Connected**  | 6         | 2/3 of platform universe                    |
| **Taxonomy AI Calls**    | 300       | Heavy classification usage                  |
| **Watermark Operations** | Unlimited | Zero marginal cost                          |
| **AI Clip Operations**   | 50        | Aligns with 50GB storage                    |
| **Thumbnail Operations** | Unlimited | Cheap operation                             |
| **Team Seats**           | 1         | Still solo operation                        |

### Studio (\$499/mo)

Full-featured tier for teams and agencies managing multiple creators.

| Feature                  | Limit     | Reasoning                                      |
| ------------------------ | --------- | ---------------------------------------------- |
| **AI-Assisted Posts**    | Unlimited | At \$499 they're effectively paying for server |
| **Posts Published**      | Unlimited | No restrictions                                |
| **Scheduled Queue Size** | Unlimited | Professional workflow support                  |
| **Platforms Connected**  | 9         | Full platform universe                         |
| **Taxonomy AI Calls**    | Unlimited | No restrictions                                |
| **Watermark Operations** | Unlimited | Zero marginal cost                             |
| **AI Clip Operations**   | Unlimited | Full video processing access                   |
| **Thumbnail Operations** | Unlimited | No restrictions                                |
| **Team Seats**           | 5         | **Anchor differentiator** for Studio tier      |

## Usage Limits JSON Schema

This configuration is stored in Directus and enforced server-side:

```json theme={null}
{
  "starter": {
    "posts_ai_assisted": 3,
    "posts_published": 5,
    "scheduled_queue_size": 3,
    "platforms_connected": 1,
    "taxonomy_ai_calls": 5,
    "watermark_ops": -1,
    "ai_clip_ops": 0,
    "thumbnail_ops": 0,
    "team_seats": 1
  },
  "creator": {
    "posts_ai_assisted": 30,
    "posts_published": 90,
    "scheduled_queue_size": -1,
    "platforms_connected": 3,
    "taxonomy_ai_calls": 75,
    "watermark_ops": -1,
    "ai_clip_ops": 10,
    "thumbnail_ops": -1,
    "team_seats": 1
  },
  "pro": {
    "posts_ai_assisted": 150,
    "posts_published": -1,
    "scheduled_queue_size": -1,
    "platforms_connected": 6,
    "taxonomy_ai_calls": 300,
    "watermark_ops": -1,
    "ai_clip_ops": 50,
    "thumbnail_ops": -1,
    "team_seats": 1
  },
  "studio": {
    "posts_ai_assisted": -1,
    "posts_published": -1,
    "scheduled_queue_size": -1,
    "platforms_connected": 9,
    "taxonomy_ai_calls": -1,
    "watermark_ops": -1,
    "ai_clip_ops": -1,
    "thumbnail_ops": -1,
    "team_seats": 5
  }
}
```

## Important Implementation Notes

<Warning>
  Review these carefully before implementing tier enforcement
</Warning>

### 1. Thumbnail Operations on Starter

`thumbnail_ops` on Starter is **0, not -1**. Thumbnails require video frame extraction. Since video is blocked on Starter (CSAM risk + cost), thumbnails must also be blocked.

**Wrong**: `"thumbnail_ops": -1` (accidentally enables video processing)
**Correct**: `"thumbnail_ops": 0` (properly blocks video operations)

### 2. Posts Published Counting

The current implementation treats **one publish event = one content piece** regardless of how many platforms it's pushed to simultaneously.

**Example**:

* User schedules 1 post to 3 platforms simultaneously = 1 count
* User schedules 3 separate posts to 1 platform = 3 counts

If you count per-platform-delivery instead, Creator at 90 posts becomes restrictive (3 platforms × 30 posts = only 30 pieces of content).

<Note>
  Document your counting convention clearly to avoid support tickets
</Note>

### 3. Team Seats on Studio

Studio tier includes **5 team seats**. You need a per-seat cost model for additional seats, or subscribers will onboard entire teams.

**Recommended overage pricing**:

* Additional seats: +\$25/seat/month
* Or: Hard cap at 5 seats, require custom enterprise pricing

## Usage Tracking

Usage is tracked in the `user_personas` collection and enforced by:

* **API endpoint**: `server/endpoints/api/usage.js`
* **Enforcement**: Each operation checks current usage vs. tier limits
* **Reset period**: Monthly, on billing anniversary
* **Storage**: `user_personas.usage_current_period` JSON field

### Usage API Example

```javascript theme={null}
const usage = await api.get('/api/usage/current');

if (usage.posts_ai_assisted >= usage.tier_limits.posts_ai_assisted) {
  throw new Error('Monthly AI post limit reached. Upgrade to continue.');
}
```

## Upgrade Flow

1. **User initiates upgrade**: From pricing page or usage limit warning
2. **Payment processing**: Payment gateway integration required
3. **Tier update**: PATCH `user_personas.pricing_tier`
4. **Usage reset**: Clear current period counters
5. **Immediate access**: New limits take effect instantly

## Downgrade Considerations

* **Graceful degradation**: Existing scheduled posts remain in queue
* **Connection preservation**: Platform connections stay active, new ones blocked
* **Data retention**: All media and history preserved
* **Immediate enforcement**: New limits apply at start of next billing period

## Related Files

* **Usage tracking**: `server/endpoints/api/usage.js`
* **Tier enforcement**: Applied in `genieChat.js`, `queue.js`, `credentials.js`
* **Frontend pricing**: `dashboard/src/pages/Pricing/index.jsx`
* **User personas**: Directus `user_personas` collection stores current tier