Mixture-of-experts model with 8x7B architecture.
mixtral-8x7b-instruct-together
LLM Gateway routes requests to the best providers that are able to handle your prompt size and parameters.