Hydra App – Credits Usage Rates
Hydra credits are bought in packages through the in-app purchase screen. The purchase price per credit depends on the package acquired. Check the stores for current price of the packages in your currency/region.
The credits purchased are loaded in the user account and are consumed on an ongoing basis as you use the app functions.
The table bellow shows the cost in credit for the different app functions, depending on the AI model selected:
Each function execution has a cost of 0.4 Hydra credits plus the applicable function cost.
Hydra credits usage on Chat functions:
Chat usage is charged based on the input tokens, output tokens and the selected chat model. The table below presents of the chat usage in Hydra credits per 1000 tokens for the different models, though the consumption is calculated per the actual number of tokens used and not rounded to 1000.
| AI Chat model selected | Hydra credits per 1000 input tokens | Hydra credits per 1000 output tokens |
|---|---|---|
| Open AI – GPT 3.5 Turbo | 0.25 | 0.75 |
| Open AI – GPT 4 | 15 | 30 |
| Open AI – GPT 4 Turbo | 5 | 15 |
| Open AI – GPT 4o | 2.5 | 7.5 |
| Open AI – GPT 4o mini | 0.075 | 0.3 |
| Open AI – o1 mini | 1.5 | 6 |
| Open AI – o1 Preview | 7.5 | 30 |
| Google – Gemini 1.5 Pro | 1.25 | 0.5 |
| Google – Gemini 1.5 Flash | 0.075 | 3 |
| Google – Gemini 1.0 Pro | 0.25 | 0.75 |
| Google – Gemini 2.0 Flash experimental | 2 | 1 |
| Mistral – Large | 1 | 3 |
| Mistral – Open 7b | 0.125 | 0.125 |
| Mistral – Open 8x7b | 0.35 | 0.35 |
| Mistral – Open 8x22b | 1 | 3 |
| Mistral – Small | 0.1 | 0.3 |
| Mistral – Nemo | 0.075 | 0.075 |
| Mistral – Codestral | 0.1 | 0.3 |
| Mistral – Pixtral 12b | 0.075 | 0.075 |
| Mistral – Pixtral Large | 1 | 3 |
| Anthropic – Claude 3 Opus | 7.5 | 37.5 |
| Anthropic – Claude 3 Sonnet | 1.5 | 7.5 |
| Anthropic – Claude 3 Haiku | 0.125 | 0.625 |
| Anthropic – Claude 3.5 Sonnet | 1.5 | 7.5 |
| Anthropic – Claude 3.5 Haiku | 0.4 | 2 |
| Meta Llama 3 – 70b | 0.295 | 0.395 |
| Meta Llama 3 – 8b | 0.12 | 0.12 |
| Meta Llama 3.1 – 405b | 1.5 | 1.5 |
| Meta Llama 3.1 – 70b | 0.45 | 0.45 |
| Meta Llama 3.1 – 8b | 0.1 | 0.1 |
| DeepSeek – R1 | 4 | 4 |
| DeepSeek – V3 | 0.45 | 0.45 |
| Qwen – 2.5 Coder 32b | 0.45 | 0.45 |
| Qwen – QWQ 32b | 0.45 | 0.45 |
| Qwen – 2.5 72b | 0.45 | 0.45 |
| Qwen – 2 VL 72b | 0.45 | 0.45 |
Tokens are the essential units that Language Models (LLMs) like ChatGPT use to understand and generate text. Tokens determine how the model processes text and how much you’re charged for its use.
What are Tokens?
Tokens are pieces of text. They can be as small as a single letter or as large as a whole word. For example, the word “unbelievable” might be split into multiple tokens: [“un”, “believ”, “able”].
A phrase like “New York” could be a single token: [“New York”].
Input Tokens
Input Tokens are the tokens you provide to the model. If you type in the chat, “How’s the weather in New York?” it might break down into 6 tokens: [“How”, “‘s”, “the”, “weather”, “in”, “New York”, “?”].
The model uses these tokens to understand your question.
Output Tokens
Output Tokens are the tokens the model generates in response. If it replies, “It’s sunny in San Francisco,” it could be tokenized as: [“It”, “‘s”, “sunny”, “in”, “San Francisco”].
These tokens form the model’s response to your input.
Hydra credits usage on Image generation functions:
When using Image generation or Image variant function the usage cost is calculated per image generated. The table below presents the amount of Hydra credits charged per each image generated, depending on the model selected.
| AI Image model selected | Hydra credits per image generated |
|---|---|
| dall-e-2@256×256 | 5100 |
| dall-e-2@512×512 | 5725 |
| dall-e-2@1024×1024 | 6350 |
| dall-e-3@1024×1024 | 12600 |
| dall-e-3@1792×1024 | 25100 |
| dall-e-3@1024×1792 | 25100 |
| dall-e-3@1024x1024hd | 25100 |
| dall-e-3@1024x1792hd | 37600 |
| dall-e-3@1792x1024hd | 37600 |
| Stability Ai Stable Ultra | 4100 |
| Stability Ai SD3 Large | 3350 |
| Stability Ai SD3 Large Turbo | 2100 |
| Stability Ai SD3 Medium | 1850 |