Black Forest Labs

Black Forest Labs is a German generative AI company headquartered in Freiburg, founded in 2024 by Robin Rombach, Patrick Esser, and Andreas Blattmann — the core researchers who created the original Stable Diffusion latent diffusion models. The company develops FLUX, a family of open-weight text-to-image and image-editing models that became the most widely adopted open-source image generation models of 2025–2026.

Its FLUX.1 release (August 2024) shipped in three tiers — FLUX.1 [pro] (commercial API), FLUX.1 [dev] (open-weight, non-commercial), and FLUX.1 [schnell] (Apache 2.0, fastest) — and FLUX.1 Kontext (2025) added in-context image editing that lets users modify images through natural-language instructions while preserving character and style consistency. In November 2025 the company released FLUX.2, its strongest model to date, supporting up to 4K-resolution generation and multi-reference conditioning across up to 10 input images for consistent characters and styles.

In December 2025 Black Forest Labs raised a $300 million Series B at a $3.25 billion post-money valuation, co-led by Salesforce Ventures and Andreessen Horowitz (a16z), with participation from NVIDIA, General Catalyst, Temasek, Bain Capital Ventures, Canva, and Figma Ventures, bringing total funding above $450 million. FLUX models power image generation inside major products including Grok (xAI), Mistral Le Chat, Adobe Firefly integrations, Canva, and Figma, and are distributed through Replicate, Fal, Together AI, and the major cloud platforms.

The company positions itself as the open-weight challenger to closed image models such as OpenAI’s GPT image generation, Google Imagen, and Midjourney, giving developers and enterprises the ability to self-host, fine-tune, and commercially license frontier image generation. Black Forest Labs competes most directly with Stability AI (founded by some of the same Stable Diffusion researchers), Midjourney, and Ideogram in the text-to-image market, and is widely regarded as the technical leader in open-weight visual generative AI.