Model Providers

In this section you will learn about the model providers available on OmniAI.

OmniAI provides universal API access to more than 30 industry leading models from OpenAI, Anthropic, Meta, Mistral, and Amazon.

Models by Provider:

OpenAI

Model	Function	Context Window
gpt-4o	Chat Completion	128,000 tokens
gpt-4-turbo	Chat Completion	128,000 tokens
gpt-3.5-turbo	Chat Completion	16,385 tokens
text-embedding-3-large	Embeddings	8,191 tokens
text-embedding-3-small	Embeddings	8,191 tokens
text-embedding-ada-002	Embeddings	8,191 tokens

Anthropic

Model	Function	Context Window
claude-3-sonnet	Chat Completion	200,000 Tokens
claude-3-haiku	Chat Completion	200,000 Tokens
claude-v2:1	Chat Completion	200,000 Tokens
claude-v2	Chat Completion	100,000 Tokens
claude-instant-v1	Chat Completion	100,000 Tokens

Meta

Model	Function	Context Window
llama3-8b-instruct	Chat Completion	8,000 Tokens
llama3-70b-instruct	Chat Completion	8,000 Tokens
llama2-13b-chat	Chat Completion	4,096 Tokens
llama2-70b-chat	Chat Completion	4,096 Tokens
llama2-13b-chat	Chat Completion	4,096 Tokens
llama2-70b-chat	Chat Completion	4,096 Tokens

Mistral

Model	Function	Context Window
mistral-7b-instruct	Chat Completion	32,000 Tokens
mixtral-8x7b-instruct	Chat Completion	32,000 Tokens
mistral-large-2402	Chat Completion	32,000 Tokens
mistral-small-2402	Chat Completion	32,000 Tokens

Amazon

Model	Function	Context Window
titan-embed-text	Embeddings	8,000 Tokens
titan-text-lite	Embeddings	4,000 Tokens
titan-text-express	Embeddings	8,000 Tokens
titan-image-generator	Image Generation	512 Characters
titan-embed-image	Image Embeddings	128 tokens
titan-text-premier	Chat Completion	32,000 Tokens
titan-embed-text	Embeddings	8,000 Tokens

Adding a Provider

Models from AWS Bedrock or OpenAI can be added in Settings. You can also add provider credentials via the Upsert Provider endpoint.