Provider Selection

How we selected the inference providers in the V1 of our leaderboard

In the V1 of our provider leaderboard, we selected providers based on popularity and SEO results. We also included certain newer, less well known providers if they excel in one criteria, such as cost. The selected provider needs to offer at least one of Llama2-70b-chat or Mixtral-8x7b, since these are the most requested, strongest open source models that our V1 will focus on. We hope to continue adding more providers and models to our leaderboard.

Since many customers default to OpenAI and Anthropic models, we included the closed source providers and their most used text-to-text models here for easy comparison.

Please contact us at [e-mail] if you wish to be included in the leaderboard as a provider.