Flux

A high-quality open-source text-to-image model from Black Forest Labs, renowned for its outstanding text rendering and human anatomy.

Flux is a series of open-source text-to-image models launched by Black Forest Labs (founded by the original Stable Diffusion team following their departure). Released in August 2024, the Flux.1 series comes in three variants: Pro (best image quality via API), Dev (open-source, cleared for commercial use), and Schnell (an ultra-fast local version) — quickly setting a new benchmark for open-source image generation.

Flux’s most prominent strengths are text rendering and human anatomy. When generating posters with text or logo scene images, the characters Flux produces are noticeably more legible than those of other open-source models from the same period, and hands and poses in portraits also look more natural. Developers can access Flux through the official Black Forest Labs API, or download the weights from HuggingFace and deploy it locally.

In terms of ecosystem positioning, Flux is the most noteworthy image generation option beyond Midjourney: for enterprises and developers that need self-hosting, customization, or cost control, being open source and locally runnable is the core appeal; for everyday users, the official website and the many third-party platforms that have integrated Flux lower the barrier to entry.

Note: Flux.1 Dev uses a hybrid license — trained on non-commercial terms but with a license that permits commercial use — so confirm the specific version’s license before deployment; running schnell locally still requires 8GB+ of VRAM; its prompting style differs from Midjourney’s and takes some tuning.

Source: Agentopia · See /llms.txt for the full site machine catalog