# SOM: Minimal Smalltalk for VM Research

> Published 2026-04-07 · https://www.promptzone.com/wiebke_chakraborty/som-minimal-smalltalk-for-vm-research-jci

Black Forest Labs has released **FLUX.2 [klein]**, a compact model series designed for real-time local image generation and editing, outperforming existing tools in speed and efficiency.


> **Model:** FLUX.2 [klein] | **Parameters:** 4B / 9B | **Speed:** 0.3-0.5s per image
> **VRAM:** 8.4 GB (4B) / 19.6 GB (9B) | **License:** Apache 2.0 (4B) / Non-commercial (9B)

## Sub-Second Generation on Consumer GPUs

The 4B variant of FLUX.2 [klein] generates **1024x1024 images in under one second**, achieving speeds 30% faster than competitors like Qwen-Image-Edit. It operates on an **RTX 4070 or 3090** with minimal setup, requiring only 8.4 GB of VRAM. The 9B model, while slightly slower at 0.5 seconds per image, enhances photorealism without sacrificing core functionality.

| Feature      | FLUX.2 klein 4B | FLUX.2 klein 9B | Qwen-Image-Edit |
|--------------|-----------------|-----------------|-----------------|
| Speed       | 0.3s           | 0.5s           | ~2s            |
| VRAM        | 8.4 GB         | 19.6 GB        | 20+ GB         |
| Editing     | Yes            | Yes            | Yes            |
| License     | Apache 2.0     | Non-commercial | Open           |

> **Bottom line:** FLUX.2 [klein] sets a new benchmark for fast, accessible image tools on consumer hardware.


![SOM: Minimal Smalltalk for VM Research](https://i.ytimg.com/vi/8GRwNM3hBDA/maxresdefault.jpg)

## Why This Matters for Local Workflows

Local AI tools like Qwen-Image require 12-16 GB of VRAM for text-to-image tasks, but integrated editing has been slower, with Qwen-Image-Edit needing over 20 GB and taking around 2 seconds per operation. FLUX.2 [klein] combines generation and editing in one model, running both under a second on standard GPUs. For developers, this means building responsive creative applications without high-end infrastructure.

Early testers on Hacker News note its potential to bridge gaps in local AI editing, with the original post earning **38 points**. This unification could accelerate workflows in fields like digital art and content creation.

{% details "Where to access" %}
- **Hugging Face:** [black-forest-labs/FLUX.2-klein](https://huggingface.co/black-forest-labs)
- **API:** Available via BFL with specific pricing tiers
- **[ComfyUI](/jaroslav/how-to-install-and-run-sdxl-models-in-comfyui-a-complete-guide-2nk2):** Community nodes for easy integration
{% enddetails %}

> **Bottom line:** This model democratizes advanced image editing for everyday developers.

In the evolving AI landscape, FLUX.2 [klein] could inspire more efficient local tools, potentially reducing reliance on cloud services and fostering innovation in real-time applications.