---
title: "Empero Shrinks a 2.4T Parameter Teacher Into a 2.78 GB Laptop Model"
subtitle: "Empero distilled a 2.4 trillion parameter teacher into a 4B model that runs on laptops with 3GB of RAM, with a big MMLU jump."
url: "https://alphasignal.ai/news/empero-shrinks-a-2-4t-parameter-teacher-into-a-2-78-gb-laptop-model"
author: "AlphaSignal Newsroom"
author_url: "https://alphasignal.ai/editorial-team"
company: "AlphaSignal"
company_url: "https://alphasignal.ai/company/alphasignal"
published: "2026-08-15T18:08:45.000Z"
updated: "2026-10-02T03:43:28.975Z"
type: "Model"
topics:
  - "Llms"
  - "Gpus"
subtopics:
  - "Small Models"
  - "Quantization"
upvotes: 156
source_url: "https://huggingface.co/empero-ai/Qwen3.8-4B-Distill-GGUF"
comments: 0
truncated: true
---

# Empero Shrinks a 2.4T Parameter Teacher Into a 2.78 GB Laptop Model

> Empero distilled a 2.4 trillion parameter teacher into a 4B model that runs on laptops with 3GB of RAM, with a big MMLU jump.

# Empero distills Qwen3.8 reasoning into a laptop-scale 4B model

Empero has released [GGUF files](https://huggingface.co/empero-ai/Qwen3.8-4B-Distill-GGUF) for Qwen3.8-4B-Distill, a reasoning model based on the Qwen3.5-4B architecture. Its smallest recommended build occupies 2.78 GB on disk, making local inference practical on many laptops and consumer desktops.

Empero says it trained the 4B student on reasoning traces from Qwen3.8 2.4T A95B, a mixture-of-experts teacher with 2.4 trillion total parameters and 95 billion active per token. The release offers developers a compact, Apache 2.0-licensed model for testing local reasoning, long-context workloads, and function calling without a hosted inference API.

## Five files, different memory bills

The repository provides five builds for llama.cpp-compatible runtimes, including Ollama, LM Studio, Jan, and KoboldCpp. Empero recommends Q4\_K\_M for general use.

__Available model files__
| Build    | File size | Intended use                                |
| -------- | --------- | ------------------------------------------- |
| Q4\_K\_M | 2.78 GB   | Recommended balance of size and quality     |
| Q5\_K\_M | 3.16 GB   | Higher fidelity with a modest size increase |
| Q6\_K    | 3.56 GB   | High-fidelity quantized inference           |
| Q8\_0    | 4.61 GB   | Largest quantized build                     |
| BF16     | 8.67 GB   | Unquantized reference weights               |

The listed sizes cover model files only. Runtime memory also includes context buffers, backend overhead, and GPU allocations. A 4 GB GPU may require a shorter context, smaller batch, or partial CPU offload, while CPU inference needs enough system RAM for the weights and working memory.

## Teacher traces shape the student

Off-policy distillation trains a smaller model against outputs produced by a larger teacher. Empero used about 45,000 curated traces covering mathematics, general reasoning, and instruction following, then updated every student parameter through full fine-tuning. The result is a standalone checkpoint rather than a LoRA adapter.

_Preview only. Full text requires an AlphaSignal Pro account._

## Sources

- [empero-ai/Qwen3.8-4B-Distill-GGUF](https://huggingface.co/empero-ai/Qwen3.8-4B-Distill-GGUF)
- [empero-ai/Qwen3.8-4B main model card](https://huggingface.co/empero-ai/Qwen3.8-4B)
- [Qwen/Qwen3.5-4B base model](https://huggingface.co/Qwen/Qwen3.5-4B)