Hermes 4 Technical Report

Ryan Teknium, Roger Jin, Jai Suphavadeeprasit, Dakota Mahan, Jeffrey Quesnelle +4 more
8/25/2025
cs.AI

No confident code implementation found for this paper.

Abstract

We present Hermes 4, a family of hybrid reasoning models that combine structured, multi-turn reasoning with broad instruction-following ability. We describe the challenges encountered during data curation, synthesis, training, and evaluation, and outline the solutions employed to address these challenges at scale. We comprehensively evaluate across mathematical reasoning, coding, knowledge, comprehension, and alignment benchmarks, and we report both quantitative performance and qualitative behavioral analysis. To support open research, all model weights are published publicly at https://huggingface.co/collections/NousResearch/hermes-4-collection-68a731bfd452e20816725728

View on arXivView PDF

Code Implementations

No confident code match yet

We couldn't find an author-owned or strongly-evidenced community implementation for this paper. Any repos shown below are weak matches — verify before relying on them.

No code implementations found yet.

Know of an implementation? Let us know in the comments below!

Similar Papers

LogiCoT: Logical Chain-of-Thought Instruction-Tuning

83% match1 repo542023

Phi-4-reasoning Technical Report

83% match2025

Mathematical Reasoning in Large Language Models: Benchmarks, Architectures, Evaluation, and Open Challenges

82% match2026

SUPERNOVA: Eliciting General Reasoning in LLMs with Reinforcement Learning on Natural Instructions

82% match2026

PRISM-MCTS: Learning from Reasoning Trajectories with Metacognitive Reflection

82% match2026

MAmmoTH: Building Math Generalist Models through Hybrid Instruction Tuning

82% match1 repo3832023

Cite this paper

@article{teknium2025hermes,
  title  = {Hermes 4 Technical Report},
  author = {Ryan Teknium and Roger Jin and Jai Suphavadeeprasit and Dakota Mahan and Jeffrey Quesnelle and Joe Li and Chen Guang and Shannon Sands and Karan Malhotra},
  year   = {2025},
  eprint = {2508.18255},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2508.18255v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2508.18255.svg)](https://codeofpaper.com/paper/2508.18255)

Discussion