# hermes-4.3-36b

Public

Fork

0

Star

2

[Use Model in LM Studio](/content/deeplink?owner=nousresearch&name=hermes-4.3-36b/index.html)

[Model](/content/nousresearch/hermes-4.3-36b ""/index.html) [Revisions](/content/nousresearch/hermes-4.3-36b/revisions ""/index.html)

Description

Hermes 4.3 36B is a frontier, hybrid-mode reasoning model based on ByteDance Seed 36B base, made by Nous Research that is aligned to you.

Stats

45 Downloads

[2 stars](/content/nousresearch/hermes-4.3-36b/stars/index.html)

Capabilities

Trained for tool use

ReasoningSupports reasoning

Minimum system memory

21GB

Tags

36B

seed\_oss

Last updated

[Updated on December 5](/content/nousresearch/hermes-4.3-36b/revisions ""/index.html) by [nous](/content/nous ""/index.html)

[Report](/content/nousresearch/hermes-4.3-36b/report ""/index.html)

README

FormattedRaw

# Hermes 4.3 - Seed 36B [Link to this section](/content/nousresearch/hermes-4.3-36b\#hermes-43---seed-36b/index.html)

## Model Description [Link to 'Model Description'](/content/nousresearch/hermes-4.3-36b\#model-description "Link to 'Model Description'"/index.html)

Hermes 4.3 36B is a frontier, hybrid-mode **reasoning** model based on ByteDance Seed 36B base, made by Nous Research that is aligned to **you**.

This is our first Hermes model trained in a decentralized manner over the internet using [Psyche](https://psyche.network/), read the blog post: [https://nousresearch.com/introducing-hermes-4-3/](https://nousresearch.com/introducing-hermes-4-3/)

Read the Hermes 4 technical report here: [Hermes 4 Technical Report](https://arxiv.org/abs/2508.18255)

Chat with Hermes in Nous Chat: [https://chat.nousresearch.com](https://chat.nousresearch.com/)

Training highlights include a newly synthesized post-training corpus emphasizing verified reasoning traces, massive improvements in math, code, STEM, logic, creativity, and format-faithful outputs, while preserving general assistant quality and broadly neutral alignment.

## What’s new vs Hermes 3 [Link to 'What’s new vs Hermes 3'](/content/nousresearch/hermes-4.3-36b\#whats-new-vs-hermes-3 "Link to 'What’s new vs Hermes 3'"/index.html)

- **Post-training corpus**: Massively increased dataset size from 1M samples and 1.2B tokens to **~5M samples / ~60B tokens** blended across reasoning and non-reasoning data.
- **Hybrid reasoning mode** with explicit `<think>…</think>` segments when the model decides to deliberate, and options to make your responses faster when you want.
- **Reasoning** that is top quality, expressive, improves math, code, STEM, logic, and even creative writing and subjective responses.
- **Schema adherence & structured outputs**: trained to produce valid JSON for given schemas and to repair malformed objects.
- **Much easier to steer and align**: extreme improvements on steerability, especially on reduced refusal rates.

## Our Mission: Frontier Capabilities Aligned to You [Link to 'Our Mission: Frontier Capabilities Aligned to You'](/content/nousresearch/hermes-4.3-36b\#our-mission-frontier-capabilities-aligned-to-you "Link to 'Our Mission: Frontier Capabilities Aligned to You'"/index.html)

In pursuit of the mission of producing models that are open, steerable and capable of producing the full range of human expression, while being able to be aligned to your values, we created a new benchmark, RefusalBench, that tests the models willingness to be helpful in a variety of scenarios commonly disallowed by closed and open models.

**Hermes 4.3 36B is now SOTA across non-abliterated models on the RefusalBench Leaderboard, surpassing our previous best of 59.5% on Hermes 4 70B**

### % of Questions Answered – RefusalBench [Link to this section](/content/nousresearch/hermes-4.3-36b\#-of-questions-answered--refusalbench/index.html)

_(Average of 5 trials)_

| Model | % of Questions Answered |
| --- | --- |
| **Hermes 4.3 36B Non-Reasoning** | 74.60% |
| **Hermes 4.3 36B Reasoning** | 72.29% |
| Hermes 4 70B Reasoning | 59.50% |
| Hermes 4 405B Reasoning | 57.10% |
| grok4 | 51.30% |
| Hermes 4 70B | 49.07% |
| Hermes 4 405B | 43.20% |
| Qwen2.5 7B | 36.10% |
| Qwen3 235B Reasoning | 34.30% |
| DeepSeek V3 | 28.10% |
| Gemini 2.5 Pro | 24.23% |
| Llama 405B | 21.70% |
| Gemini 2.5 Flash | 19.13% |
| GPT4o | 17.67% |
| Sonnet 4 | 17.00% |
| GPT4-mini | 16.76% |
| R1 | 16.70% |
| cogito-v2-405B Reasoning | 15.40% |
| Opus 4.1 | 15.38% |
| Qwen3 235B | 15.30% |
| cogito-v2-405B | 14.94% |
| cogito-v2-405B | 12.10% |
| GPT 5 | 11.34% |
| gpt-oss 120B | 5.60% |
| gpt-oss 20B | 4.79% |

Hermes 4 achieves SOTA on RefusalBench across all popular closed and open models in being helpful and conforming to your values, without censorship.

## Benchmarks (Hermes 4.3 36B) [Link to 'Benchmarks (Hermes 4.3 36B)'](/content/nousresearch/hermes-4.3-36b\#benchmarks-hermes-43-36b "Link to 'Benchmarks (Hermes 4.3 36B/index.html)'")

|  | Hermes 4.3 36B Psyche | Hermes 4.3 36B Centralized | Hermes 4 70B Centralized |
| --- | --- | --- | --- |
| AIME 24 | 71.9 | 70.6 | 73.5 |
| AIME 25 | 69.3 | 66.8 | 67.4 |
| BBH | 86.4 | 84.7 | 87.8 |
| DROP | 83.5 | 81.6 | 85.0 |
| GPQA Diamond | 65.5 | 64.8 | 66.1 |
| IFEval | 77.9 | 73.9 | 78.7 |
| MATH-500 | 93.8 | 92.3 | 95.5 |
| MMLU | 87.7 | 86.5 | 88.4 |
| MMLU-Pro | 80.7 | 79.7 | 80.7 |
| MuSR | 69.7 | 64.7 | 70.4 |
| OBQA | 96.6 | 91.8 | 94.8 |
| SimpleQA | 6.0 | 5.6 | 17.9 |

## Prompt Format [Link to 'Prompt Format'](/content/nousresearch/hermes-4.3-36b\#prompt-format "Link to 'Prompt Format'"/index.html)

Hermes 4 uses Llama-3-Chat format with role headers and special tags.

**Basic chat:**

```
<|start_header_id|>system<|end_header_id|>

You are Hermes 4. Be concise and helpful.<|eot_id|>
<|start_header_id|>user<|end_header_id|>

Explain the photoelectric effect simply.<|eot_id|>
<|start_header_id|>assistant<|end_header_id|>
```

### Reasoning mode [Link to this section](/content/nousresearch/hermes-4.3-36b\#reasoning-mode/index.html)

Reasoning mode can be activated with the chat template via the flag `thinking=True` or by using the following system prompt:

```
You are a deep thinking AI, you may use extremely long chains of thought to deeply consider the problem and deliberate with yourself via systematic reasoning processes to help come to a correct solution prior to answering. You should enclose your thoughts and internal monologue inside <think> </think> tags, and then provide your solution or response to the problem.
```

Note that you can add any additional system instructions before or after this system message, and it will adjust the models policies, style, and effort of thinking, as well as its post-thinking style, format, identity, and more. You may also interleave the tool definition system message with the reasoning one.

When the model chooses to deliberate, it emits:

```
<|start_header_id|>assistant<|end_header_id|>
<think>
…model’s internal reasoning may appear here…
</think>
Final response starts here…<|eot_id|>
```

Additionally, we provide a flag to keep the content inbetween the `<think> ... </think>` that you can play with by setting `keep_cots=True`

## Function Calling & Tool Use [Link to 'Function Calling & Tool Use'](/content/nousresearch/hermes-4.3-36b\#function-calling--tool-use "Link to 'Function Calling & Tool Use'"/index.html)

Hermes 4 supports function/tool calls _within_ a single assistant turn, produced after it's reasoning:

**System message (example):**

```
<|start_header_id|>system<|end_header_id|>
You are a function-calling AI. Tools are provided inside <tools>…</tools>.
When appropriate, call a tool by emitting a <tool_call>{...}</tool_call> object.
After a tool responds (as <tool_response>), continue reasoning inside <think> and produce the final answer.
<tools>
{"type":"function","function":{"name":"get_weather","description":"Get weather by city","parameters":{"type":"object","properties":{"city":{"type":"string"}},"required":["city"]}}}
</tools><|eot_id|>
```

Note that you may also simply place tool definitions into the "tools:" field of your messages, and the chat template will parse and create the system prompt for you. This also works with reasoning mode for improved accuracy of tool use.

The model will then generate tool calls within `<tool_call> {tool_call} </tool_call>` tags, for easy parsing. The tool\_call tags are also added tokens, so it makes it easy to parse while streaming! There are also automatic tool parsers built-in to VLLM and SGLang for Hermes, just set the tool parser in VLLM to `hermes` and in SGLang to `qwen25`.

## Inference Notes [Link to 'Inference Notes'](/content/nousresearch/hermes-4.3-36b\#inference-notes "Link to 'Inference Notes'"/index.html)

- **Sampling defaults that work well:**`temperature=0.6, top_p=0.95, top_k=20`.
- **Template:** Use the Llama chat format for Hermes 4.3 36B, 70B, and 405B as shown above, or set `add_generation_prompt=True` when using `tokenizer.apply_chat_template(...)`.

### Transformers example [Link to this section](/content/nousresearch/hermes-4.3-36b\#transformers-example/index.html)

```
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

model_id = "NousResearch/Hermes-4.3-36B"

tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.float16,
    device_map="auto"
)

messages = [\
    {"role":"system","content":"You are Hermes 4. Be concise."},\
    {"role":"user","content":"Summarize CRISPR in 3 sentences."}\
]

inputs = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True, return_tensors="pt"
).to(model.device)

outputs = model.generate(
    **inputs, max_new_tokens=400, temperature=0.6, top_p=0.95, top_k=20, do_sample=True
)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```

For production serving on multi-GPU nodes, consider tensor parallel inference engines (e.g., SGLang/vLLM backends) with prefix caching.

## Inference Providers: [Link to 'Inference Providers:'](/content/nousresearch/hermes-4.3-36b\#inference-providers "Link to 'Inference Providers:'"/index.html)

### Nous Portal: [Link to this section](/content/nousresearch/hermes-4.3-36b\#nous-portal/index.html)

### Chutes: [Link to this section](/content/nousresearch/hermes-4.3-36b\#chutes/index.html)

# Quantized / Smaller Variants [Link to this section](/content/nousresearch/hermes-4.3-36b\#quantized--smaller-variants/index.html)

Hermes 4 is available as BF16 original weights as well as BF16 as well as FP8 variants and GGUF variants by LM Studio.

GGUF Verions (4, 5, 6, and 8bit): [https://huggingface.co/NousResearch/Hermes-4.3-36B-GGUF](https://huggingface.co/NousResearch/Hermes-4.3-36B-GGUF)

See the Hermes 4 collection to explore them all:
[https://huggingface.co/collections/NousResearch/hermes-4-collection-68a731bfd452e20816725728](https://huggingface.co/collections/NousResearch/hermes-4-collection-68a731bfd452e20816725728)

# How to cite [Link to this section](/content/nousresearch/hermes-4.3-36b\#how-to-cite/index.html)

```
@misc{teknium2025hermes4technicalreport,
      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},
      primaryClass={cs.AI},
      url={https://arxiv.org/abs/2508.18255},
}
```

Custom Fields

Special features defined by the model author

Enable Thinking

: boolean

(default=false)

Controls whether the model will think before replying

Keep CoT

: boolean

(default=false)

Include Chain of Thought in subsequent requests

Sources

The underlying model files this model uses

Based on

🤗 [NousResearch/Hermes-4.3-36B-GGUF](https://huggingface.co/NousResearch/Hermes-4.3-36B-GGUF) →

GGUF

nousresearch/hermes-4.3-36b • LM Studio Hub

|
|

|
|

|
|