Custom models Introduction

What is a custom model?

Tinq.ai's custom models feature empowers users to train their own AI on their unique dataset, be it web articles, PDFs, or any other text content.

This allows users to have complete control over the type and quality of content generated by the AI. With the ability to train the AI on their own data, users can fine-tune the AI's understanding of their specific domain or industry, leading to more accurate and relevant content creation.

This feature is particularly beneficial for businesses and organizations that require specific language or tone of voice in their content. Tinq.ai's custom models offer users the flexibility to tailor the AI's output to their unique needs, ultimately resulting in more efficient and effective content creation.

🚧

Beta

As of now, Custom Models are in Beta, and available to all users starting at Tinq.ai's Pro plan.

Pricing

Please note that this pricing is tentative during the beta phase and is subject to change.

Pricing examples

• Training:
  • If, for example you are adding a document that has 3500 words for training, then 3500/400 = 8.75, therefore this training will consume 9 credits from your available credits.
• Querying:
  • When querying, if for example your input makes 50 words, and the output is 400 words, the total will be 450 words. 450/400 = 1.125, therefore 2 credits will be consumed from your available credits.

Please note that other factors can enter into consideration that may increase the cost of the request such as the complexity if the training data that you are providing.

Typically scientific content (papers, formula, etc...) tend to make complexity increase.

FAQ

• Is there an API for custom models? Yes, an API is available for both training & querying. Please contact us for a documentation.
• Is the pricing for the API the same as the web tool? Yes.

Please feel free to reach out to us if you have any question or need further clarification.