Sentient AI
Information
Sentient AI Models
Sentient Offers variety of State of the art AI models with lowest latency and highest inference power in the industry
Models are the backbone of our API, delivering user-related data tailored to each endpoint's unique requirements. Every endpoint utilizes specialized models, each optimized for its specific capabilities, ensuring efficient and intelligent responses.
Language models excel in tasks such as conversation, code generation, and summarization, while image generation models push the boundaries of visual creation, producing highly accurate and detailed images like never before.
Language models excel in tasks such as conversation, code generation, and summarization, while image generation models push the boundaries of visual creation, producing highly accurate and detailed images like never before.
AI Flow Diagram with Model's Integration/Importance
The model is very first component of main API Logic, once the request passes through Guard model and processed/verified by Middelware it is passed to out Core Model that will be pulled from the request's JSON structure with the following model: '<MODEL NAME>';.
The Guard model is the first model in the pipeline, as it is used to check the status of prompt whether if it is safe or not. If the prompt fails the guard model, the prompt is rejected and a response of 400 is sent to the user.
For each endpoint we have a specific model that is used to generate the response. The model is trained on a specific dataset and is optimized for the specific task that the endpoint is designed for. Using one model for a non-supported endpoint/task will result in a 400 error.
The Guard model is the first model in the pipeline, as it is used to check the status of prompt whether if it is safe or not. If the prompt fails the guard model, the prompt is rejected and a response of 400 is sent to the user.
For each endpoint we have a specific model that is used to generate the response. The model is trained on a specific dataset and is optimized for the specific task that the endpoint is designed for. Using one model for a non-supported endpoint/task will result in a 400 error.
Models
Sentient currently offers language and image gen models limited to chat and generating images, We plan to add embedding, web search and text to speech models in the future.
Language Models
Language models are generally used to generate text, chat, code, and summarization. Sentient offers a variety of language models that are optimized for each task.Image Generation Models
Image generation models are used to generate images. Sentient offers a variety of image generation models that are optimized for each task.PreviousIntroduction
NextLanguage Models
On this Page
Model Introduction
Model Integration
Models
Language Models
Image Generation Models