Microsoft ships Flux 3 — a new image model that actually works

Microsoft dropped Flux 3, a new image generation model that the team says lands at the top of its class for prompt accuracy. The model was trained on a big chunk of the open web — roughly 11 billion images and captions — and it's open weight, meaning you can run it yourself without paying a per-image API fee. The weights are available under the Apache 2 license.

What makes Flux 3 worth paying attention to is the balance it strikes: it's fast enough for production use, handles long prompts without losing the plot, and doesn't require a datacenter-grade GPU to run. On a single consumer card you can generate images in seconds. That matters for the folks building tools on top of it — the indie devs, the small shops, the people who can't afford to rent a cluster just to make a logo.

Why this matters for us: open weight means our community isn't locked into one vendor's billing. We can host it, we can fine-tune it, and when Microsoft updates it we can update it too.