# Models & Credits

## Holo Credits

Every video component (like voice, image, or scene) costs credits to generate. Here’s a rough guide:

| **Feature**                 | **API Cost (USD)** | **Base Cost in credits** | **Priced with 20% Overhead** | <p><strong>With 20% Overhead</strong><br><br><strong>(Credit)</strong></p> |
| --------------------------- | ------------------ | ------------------------ | ---------------------------- | -------------------------------------------------------------------------- |
| Video (per sec)             | $0.35              | 35                       | $0.42                        | 42 credits                                                                 |
| TTS Voice (per min)         | $0.24              | 24                       | $0.29                        | 29 credits                                                                 |
| AI Image (per image)        | $0.03              | 3                        | $0.04                        | 4 credits                                                                  |
| Stock (Pexels) ( per image) | Free               | 0                        | Free                         | 2 credits                                                                  |
| Sound FX (per SFX)          | $0–0.02            | 0–1                      | $0–0.01                      | 1 credits                                                                  |

*Credit costs are based on underlying model/API usage plus system overhead for rendering, storage, and performance.*

## Model & Providers

Below is an overview of the models and APIs we use to generate each component, along with their intended use.

| **Provider**   | **Model/API**     | **Used For**                | **Notes**                                                          |
| -------------- | ----------------- | --------------------------- | ------------------------------------------------------------------ |
| **Google**     | Veo               | Video generation            | Outputs high-res video clips (5–8s), length not fully controllable |
|                | Imagen            | Photorealistic image gen    | Great for realistic scenes or backdrops                            |
|                | Gemini Flash      | Stylized/cartoonish images  | Ideal for meme, animation, or brainrot content                     |
| **OpenAI**     | Open AI Image gen | General-purpose image gen   | Balanced option between realism and stylization                    |
| **ElevenLabs** | Multilingual TTS  | Voiceovers (text to speech) | Used across all templates for natural speech in multiple languages |
|                | Sound Effect API  | Short audio clips           | Enhances pacing with pings, wooshes, etc.                          |

We use dynamic routing based on content type and quality needs : for example, Studio will auto-select between Imagen and Gemini Flash depending on whether your prompt leans photorealistic or stylized.