January 31, 2024

Code Llama 70B now available on OctoAI.

  • Code Llama 70B Instruct now available via API or in the web UI.
    • Use the same API and token key, simply replace the models with ‘codellama-70b-instruct-fp16’

January 4, 2024

OctoAI launched a private preview for users who want to “bring your own” fine-tuned LLM.

  • Host any fine-tuned Llama 2, Code Llama, Mixtral, or Mistral models.
  • Run your fine-tuned model at the same cost and latency as the default models already on OctoAI.