AI Providers

ServerAssistantAI supports a wide range of AI providers for both language models (LLMs) and embeddings.

What Are Providers?

Providers are configurable components that enable ServerAssistantAI's customizability and flexibility. They serve as the backbone of the plugin's ability to integrate with various AI services, giving server owners the power to tailor the AI assistant's capabilities to their specific needs.

ServerAssistantAI offers several ways to configure and extend its functionality through a flexible provider system for embedding models, chat models (LLMs), and question detection, including Built-in Providers (ready-to-use), Addon Providers (additional providers installed via addons), Pre-configured OpenAI-compatible Providers (built-in providers with predefined endpoint URLs), and Custom Providers with Custom Base URLs (integration with any OpenAI API-compatible service using a custom endpoint URL).

To configure a provider, simply specify its name and any required options in the config.yml file using the following format:

section_name:
    provider: 'openai' # Example provider
    option1: value

Each provider has its own set of options. Some options are shared across providers, while others may have the same name but behave differently depending on the provider. The specific functionality and configuration are determined by the selected provider.

Supported AI Providers

Provider	Type	Functionality	Pricing	Description
Cohere		LLM & Embedding	Free & Paid	Provides access to Cohere's language and embedding models, with RAG capabilities to improve performance.
OpenAI		LLM & Embedding	Paid	Offers premium paid models like GPT-3.5-turbo and GPT-4.
Anthropic		LLM	Paid	Enables the use of Anthropic's Claude models for LLM functionality.
Azure-OpenAI		LLM & Embedding	Paid	Allows integration with Azure OpenAI Service for both LLM and embedding capabilities.
DeepInfra		LLM & Embedding	Paid	Run the latest ML models with ease using DeepInfra's simple REST API.
FireworksAI		LLM & Embedding	Paid	Fireworks.ai is a fast inference platform for serving generative AI models efficiently.
Github-Models		LLM	Free	Access industry-leading AI models directly on GitHub for free.
Google-AIStudio		LLM & Embedding	Free & Paid	Provides access to Google's most advanced Gemini generative AI models.
Groq		LLM	Temporarily Free	Utilizes Groq's LPU (Language Processing Unit) Inference Engine for fast LLM inference.
HuggingFace		LLM & Embedding	Free	Provides access to thousands of open-source models for free through the HuggingFace Inference API.
Kolank		LLM	Paid	Kolank is an AI routing platform that connects to various models, ensuring high-quality responses.
LM-Studio		LLM & Embedding	Self-hosted	A desktop app for running local models on your computer, supporting models from HuggingFace.
LocalAI		LLM & Embedding	Self-hosted	Open-source, OpenAI drop-in alternative REST API for local inferencing without a GPU.
Mistral-AI		LLM & Embedding	Paid	Integrates Mistral AI models for both LLM and embedding capabilities.
NagaAI		LLM & Embedding	Free & Paid	Offers reliable API access to GPT-4, Gemini 1.5, Llama 3B, and various other language and embedding models.
Nvidia-Models		LLM	Free & Paid	Integrates NVIDIA's optimized AI models for efficient LLM functionality.
OctoAI		LLM	Paid	Harness the latest AI innovations with OctoAI's efficient, reliable, and customizable AI systems for your apps.
Ollama		LLM	Self-hosted	Allows self-hosting of Ollama, a lightweight framework for running language models locally.
OpenLLM		LLM	Self-hosted	Allows developers to run any open-source LLMs (Llama 3.1, Qwen2, Phi3 and more) or custom models.
Openrouter		LLM	Free & Paid	Standardized API for switching between models and providers, prioritizing price or performance.
Perplexity		LLM	Paid	Perplexity AI's API enables users to use Perplexity Models and Open-Source LLMs.
TogetherAI		LLM & Embedding	Paid	Fast, cost-efficient, and scalable inference for open-source models like Llama-3.
01.AI		LLM	Paid	Offers language models like Yi-1.5, delivering strong performance in instruction-following.
xAI		LLM & Embedding	Free & Paid	Brings xAI's powerful Grok models, enabling advanced text capabilities.
OpenAI-Variant		LLM &/or Embedding	Free or Paid	Allows integration with any OpenAI API-compatible service. Users can set up custom endpoints by specifying the base URL in the config.yml file.

All providers that are not built-in or OpenAI variants require the installation of their respective addons, which are available for free.

With support for a diverse range of AI providers, ServerAssistantAI enables users to choose the models and services that best fit their needs and budget.