Supported LLM providers

Supported LLM Providers

The CAST AI LLM Proxy integrates with various Large Language Model (LLM) providers, enabling efficient request routing based on complexity and cost. This document outlines the supported providers and their available models for routing and proxying.

Provider Overview

Below is a comprehensive table of supported providers, their API identifiers, any additional required fields, and the models they support:

ProviderAPI IdentifierAdditional Required FieldsRoutable and Proxyable Models
OpenAIOPENAIN/Agpt-4-0125-preview, gpt-4-0613, gpt-4-1106-preview, gpt-4o-2024-05-13, gpt-4-turbo-2024-04-09
AzureAZUREurl and apiVersioncommand-r-plus, gpt-4-0613, gpt-4-1106-preview, gpt-4-turbo-2024-04-09, mistral-large-2402
AzureAIAZUREAIurlmistral-small
AnthropicANTHROPICN/Aclaude-2.1, claude-3-5-sonnet-20240620, claude-3-haiku-20240307, claude-3-opus-20240229, claude-3-sonnet-20240229
GeminiGEMININ/Agemini-1.0-pro, gemini-1.5-flash, gemini-1.5-pro
VertexAI (Gemini)VERTEXAIGEMINIurlgemini-1.0-pro, gemini-1.5-flash, gemini-1.5-pro
VertexAI (Anthropic)VERTEXAIANTHROPICurlclaude-3-5-sonnet-20240620, claude-3-haiku-20240307, claude-3-opus-20240229, claude-3-sonnet-20240229
GroqGROQN/Agemma-7b-it, llama3-70b-8192, llama3-8b-8192, mixtral-8x7b-32768
MistralMISTRALN/Acodestral-2405, mistral-large-2402, mistral-medium-2312, mistral-small, mixtral-8x7b-32768, open-mistral-7b, open-mixtral-8x22b
OllamaOLLAMAurlllama3-70b-8192, llama3-8b-8192, llama2:7b, llama2:70b
CohereCOHEREN/Acommand-light, command-r, command-r-plus
DatabricksDATABRICKSurldatabricks-dbrx-instruct
CodestralCODESTRALN/Acodestral-2405
OpenRouterOPENROUTERN/Aclaude-3-haiku-20240307, command-r-plus, gpt-4o-2024-05-13
AnyscaleANYSCALEN/Agemma-7b-it

This table provides a quick reference for the supported providers, their API identifiers, any additional fields required for configuration, and the models available for routing and proxying. For detailed information on configuring and using these providers with the CAST AI LLM Proxy, please refer to the getting started guide.