Overview

Hermes 3 is a general-purpose fine-tune of Llama 3.1 (8B, 70B and 405B) from the open-source lab Nous Research, popular with developers who want a more steerable model.

Highlights

  • 8B, 70B and 405B (plus a 3B Llama 3.2 version)
  • Function calling and JSON output
  • Strong system-prompt adherence
  • Based on Llama 3.1

Typical uses

Agents, role-play and applications that need precise steering.

Things to consider

  • Llama license terms apply.
  • Download sizes refer to the default quantized versions in the Ollama library; full-precision weights are larger.

Specifications are taken from the model card on Hugging Face, the Ollama library and the provider’s announcement. Benchmark results depend on evaluation settings; see how AI benchmarks work.