FLUX.2 [dev]

BFL's frontier open-weights T2I — best-in-class prompt adherence and text rendering for any license-flexible open model in April 2026.

License: FLUX.2-dev Non-Commercial License — flag loudly · Context: Up to 4MP; ~28–50 steps; multi-reference image editing · Released: November 25, 2025

The decision in five lines

The call: Buy — for image
Best for: image
Runs on: 23 hardware picks fit (cheapest: Intel Arc B580 12 GB · $249)
Watch out: Any commercial product or paid output — non-commercial license is strict.
Evidence: Estimated · last verified April 2026

32B: PARAMETERS
IMAGE GEN: TYPE
Up: CONTEXT
~8–10 GB (Q4_K_M GGUF) / ~16 GB (Q8) / >80 GB (BF16): VRAM AT Q4

Where we recommend this

Every tier slot in the planner where this model is a top or alternate pick. Pulled live from planner.js — when the planner refreshes, this table stays current.

IMAGE · TOP

FLUX.2 dev (non-commercial)Quality ceiling at 32B; FLUX Non-Commercial license — commercial deployment needs paid BFL license.

The call

BFL's frontier open-weights T2I — best-in-class prompt adherence and text rendering for any license-flexible open model in April 2026.
When not to use: Any commercial product or paid output — non-commercial license is strict. Ship products via FLUX.2 Pro API or use FLUX.2 klein 4B (Apache 2.0) instead.

Runner notes

ComfyUI with GGUF quant is the usable path. No Ollama (image models aren't served there).

License: FLUX.2-dev Non-Commercial License — flag loudly
Released: November 25, 2025
Maker: Black Forest Labs
Model card: huggingface.co/black-forest-labs/FLUX.2-dev →

Hardware that fits

Every hardware pick whose memory fits this model at the quant we recommend. Sorted cheapest-first — the top row is your best-value fit. Click through for the full buyer’s guide.

Next step

Find-by-model — see what hardware runs this→