Skip to content
OmniRail
← All models

Nous · Language model

Hermes 4 405B

nousresearch/hermes-4-405b
  • Language
  • Uncensored · fewer refusals
Released Aug 26, 2025

In the Uncensored models category. Nous Research's hybrid-reasoning Llama 3.1 405B, trained to be steerable and neutral. An ordinary chat model: same endpoint, key and MCP tool as any other. These models refuse less often, but our usage policy still applies to every call: nothing illegal, nothing sexual involving minors, no harassment or threats against real people, no malware or weapons meant to hurt anyone. How it works · Usage policy

Hermes 4 is a large-scale reasoning model built on Meta-Llama-3.1-405B and released by Nous Research. It introduces a hybrid reasoning mode, where the model can choose to deliberate internally with...

Specifications

Context window
131,072 tokens
Max output
117,964 tokens
Input
text
Output
text
Reasoning
supported
Tokenizer
Other
Knowledge cutoff
2024-08-31

Supported parameters

Accepted by the model and forwarded as-is by the gateway.

  • frequency_penalty
  • logit_bias
  • logprobs
  • max_tokens
  • presence_penalty
  • reasoning
  • repetition_penalty
  • response_format
  • seed
  • stop
  • structured_outputs
  • temperature
  • top_k
  • top_logprobs
  • top_p

Use it

POST /api/v1/chat/completions · MCP chat · full reference

curl https://omnirail.org/api/v1/chat/completions \
  -H "Authorization: Bearer $OMNIRAIL_KEY" -H "Content-Type: application/json" \
  -d '{"model":"nousresearch/hermes-4-405b","messages":[{"role":"user","content":"Hello"}],"max_tokens":300}'