FLUX 3 Image Klein (5.6B, open weights) lands in ComfyUI with native 4K
crystal_alpine · reddit · 2026-10-16
Black Forest Labs released FLUX 3 Image Klein, the 5.6B compact open-weights model in the FLUX 3 family, natively supported in ComfyUI with two checkpoints: full undistilled Base for fine-tuning/LoRA/research, and a Distilled low-latency turbo variant.
Key features:
- Native 4K generation (not upscaled)
- Bounding-box layout control before generation
- Multi-reference generation/editing with subject consistency
- Text rendering plus editing of existing in-image text
- Reference-based style transfer
- A separate prompt upsampler model
Speed on consumer GPUs: INT8 ConvRot quantization + the step-distilled model runs 1024×1024 text-to-image 4.3×–9.3× faster than the BF16 50-step baseline — RTX 5070: 71s → under 8s; RTX 4070: 78s → under 9s.
Related event: Black Forest Labs Open-Sources FLUX.3 Image Klein(34 posts)→
More from Infra
- 14 open-source LLM observability tools to know in 2026, from Langfuse to Laminar — TheTuringPost · 2026-10-16
- Local AI is a routing problem, not a privacy feature: where each task should run — bigdata · 2026-10-16
- Moving agent work from GPUs to CPUs for durable, predictable actions — cramforce · 2026-10-16
- First TPUv7 agentic benchmark on vLLM claims 1.42x perf-per-dollar over GB300 NVL72 — vllm_project · 2026-10-16
- Phison CEO calls CXMT memory scare a 'ghost story', warns of materials shortage — BenBajarin · 2026-10-16
- Ben Bajarin: The Future of Compute Is Fungible as Hyperscalers Design for Workload Flexibility — BenBajarin · 2026-10-16