I Ran FLUX 3 Image Klein on My RTX 4070 — Here's What Actually Happened
Description and use
The INT8 version of FLUX 3 Image Klein runs on an RTX 4070, producing 1 MP images in under 9 s, enabling local generation for small teams.
Problem
Running a large 5.6B parameter image generation model locally on consumer hardware was uncertain and potentially slow.
Outcome
The INT8 version runs efficiently on an RTX 4070, producing 1 MP images in <9 s, proving local inference viable for small teams and indie projects.
Workflow
- 1. Install FLUX 3 Image Klein INT8 build on RTX 4070
- 2. Load model in ComfyUI
- 3. Generate a 1 MP image
- 4. Measure generation time (under 9 s)
Evidence
The INT8 build fits in 8GB VRAM and generates a 1MP image in under 9 seconds through ComfyUI.
The model is open weights under the FLUX Kommunity License, which means free commercial use if you're under $5M revenue.
Combined with LoRA fine-tuning support from day one, this changes what's realistic for small teams or indie projects.
- Record date: 2026-10-17T00:40:57.633Z