GraySoft
Projects Models Compare Cloud benchmarks FAQ Download guIDE →
Model Intelligence Sheet

RaspizdAI/petuh-2.0-gguf overview

🐔 Petuh 2.0 The fastest lightweight reasoning model trained from scratch on 4.72B tokens and fine tuned on pure entropy. Technical Specifications: Total Param…

ggufpetuhreasoningthinklicense:mitendpoints_compatibleregion:usconversational

Runs locally from ~52.1 MB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).

Downloads
0
Likes
1
Pipeline
Author

Repository Files & Downloads

8 GGUF files detected
Direct downloads for local inference
FileTypeQuantizationSizeLink
petuh-2.0-f16.ggufGGUFF16208.3 MBDownload
petuh-2.0-q3_k_m.ggufGGUFQ3_K_M52.1 MBDownload
petuh-2.0-q4_0.ggufGGUFQ4_060.5 MBDownload
petuh-2.0-q4_k_m.ggufGGUFQ4_K_M63.4 MBDownload
petuh-2.0-q5_0.ggufGGUFQ5_072.8 MBDownload
petuh-2.0-q5_k_m.ggufGGUFQ5_K_M74.3 MBDownload
petuh-2.0-q6_k.ggufGGUFQ6_K85.8 MBDownload
petuh-2.0-q8_0.ggufGGUFQ8_0110.9 MBDownload

Model Details

Model IDRaspizdAI/petuh-2.0-gguf
AuthorRaspizdAI
Pipeline
Licensemit
Base model
Last modified2026-09-08T18:32:02.000Z

Model README

---

license: mit

tags:

  • petuh
  • reasoning
  • think

---

🐔 Petuh-2.0

The fastest lightweight reasoning model trained from scratch on 4.72B tokens and fine-tuned on pure entropy.

Technical Specifications:

  • Total Parameters: 108,943,104 (~108.9M)
  • Vocabulary Size: 8,192
  • Embedding Dimensions: 768
  • Hidden Layers: 14
  • Attention Heads: 12 (Query) / 4 (KV, GQA)
  • Context Length: 1,536

Benchmark Results:

| Benchmark | Score |

| :--- | :---: |

| AIME 2025 | 0.0 |

| HLE | 0.0 |

| Terminal bench 2.1 | 0.0 |

| SWE Bench verified | 0.0 |

Run RaspizdAI/petuh-2.0-gguf with guIDE

Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.

Download guIDE → · Browse 524k+ models · Compare models

Source: Hugging Face · Compare models