Black Forest Labs Launches FLUX.1 Kontext

Black Forest Labs today unveiled FLUX.1 Kontext, a groundbreaking suite of generative flow-matching models that unites image generation and editing in a single architecture. Announced from Freiburg, Germany on May 29, 2025, FLUX.1 Kontext empowers creators, developers, and enterprises to generate, retouch, and iteratively refine images using both text and visual inputs—without any finetuning or complex workflows .

Key Features of FLUX.1 Kontext

• In-Context Generation & Editing: Kontext understands and extracts visual concepts from both prompts and reference images. Users can apply localized edits—such as changing colors, modifying text, or restyling scenes—by simply issuing natural-language instructions. This tight integration of generation and editing delivers seamless visual storytelling and rapid ideation at inference speeds up to 8× faster than leading alternatives .
• Interactive Multi-Turn Refinement: With FLUX.1 Kontext [Pro], users can iteratively build upon previous edits while preserving character consistency and style across multiple scenes and viewpoints. Edits typically complete in 3–5 seconds for 1 MP outputs, enabling truly fluid creative workflows .

Model Variants

• FLUX.1 Kontext [Pro]: Optimized for rapid, multi-step editing, Pro maintains high fidelity to both initial and subsequent prompts, ensuring character and stylistic coherence across iterations.
• FLUX.1 Kontext [Max]: Delivers maximum prompt adherence, advanced typography handling, and premium rendering quality—all at minimal latency.
• FLUX.1 Kontext [Dev]: An open-weight 12 B diffusion transformer released in private beta for research and safety testing. Ideal for customization and community-driven extensions .

Flux.1 Kontex [pro] and Flux.1 Kontex [max]. The former allows users to generate images and refine them through multiple “rotations” while preserving the character and style in the image. Flux.1 Kontex [max] focuses on speed, consistency, and following prompts.

Unlike some of Black Forest Labs’ previous models, Flux.1 Kontex [pro] and Flux.1 Kontex [max] cannot be downloaded for offline use. However, Black Forest Labs is developing an “open” Kontex model, Flux.1 Kontex [dev], and offers a closed beta version for research and safety testing.

Availability & Integrations of FLUX.1 Kontext

FLUX.1 Kontext is accessible through the BFL API and directly via the newly launched BFL Playground web interface, which grants new users 200 free credits (approximately 12 pro-model generations) without requiring third-party sign-up.

“FLUX.1 Kontext represents a fundamental shift from traditional editing approaches by unifying image generation and editing in a single flow matching architecture,” said Robin Rombach, CEO and co-founder of Black Forest Labs. “We have achieved state-of-the-art character consistency across multi-turn edits while maintaining interactive inference speeds, enabling creative workflows that were previously impossible.”

Evaluation Results

In extensive benchmarks on the proprietary KontextBench suite of six real-world in-context tasks, FLUX.1 Kontext [Pro] and [Max]:

• Top Text Editing & Character Preservation: Highest scores across both tasks, outperforming peer flow- and diffusion-based models.
• Inference Speed: Achieved up to 8× faster generation and editing cycles compared to leading alternatives (e.g., GPT-Image).
• Quality Dimensions: Demonstrated competitive performance on FID and CLIP-based realism measures, as well as human-rated aesthetics, prompt following, and typography benchmarks .

Enterprise Playground

To facilitate evaluation, Black Forest Labs has launched the BFL Playground, a browser-based demo environment where business users can test FLUX.1 Kontext’s capabilities in real time. In the coming weeks, the company will roll out additional components—such as an open-weights, distilled variant and self-deployment options with transparent pricing and commercial use rights—to broaden adoption and support enterprise integration.

As generative AI continues to reshape digital media, FLUX.1 Kontext stands out by marrying generation and editing into a cohesive, high-performance system—paving the way for more intuitive, efficient, and expressive creative workflows.

CometAPI: FLUX.1 Kontext by Black Forest Labs Coming Soon

CometAPI provides a unified REST interface that aggregates hundreds of AI models—including flux family—under a consistent endpoint, with built-in API-key management, usage quotas, and billing dashboards. Instead of juggling multiple vendor URLs and credentials.

While we finalize FLUX.1 Kontext upload, explore our other models on the Models page or try them in the AI Playground.

The FLUX.1 Kontext API will soon appear on CometAPI, so stay tuned！

While Waiting, Developers can access Midjourney API and FLUX.1 [dev]  through CometAPI to generate Video. To begin, explore the model’s capabilities in the Playground and consult the API guide for detailed instructions. Before accessing, please make sure you have logged in to CometAPI and obtained the API key.