Six Times Speed Improvement! Black Forest Labs Releases Flux 1.1 Pro: Dual Upgrade in Speed and Image Quality!

AIbase基地

Published inAI News · 4 min read · Oct 8, 2024

444

Recently, Black Forest Labs, a German AI startup, has introduced its latest image generation model, Flux1.1Pro. The speed of this new model is truly astonishing, with Black Forest Labs claiming that it is six times faster than its predecessor, Flux1Pro, while also significantly enhancing image quality.

In addition to Flux1.1Pro, Black Forest Labs has also released the beta version of its BFL API, allowing developers to integrate the Flux model into their own applications.

Flux1.1Pro has been enhanced in several key areas, not only achieving unprecedented speed levels but also improving image quality, command accuracy, and image diversity. In the future, this model will support the generation of 2K resolution images at high speed, which is highly anticipated.

In comparative tests with competitors such as Ideogram v2 and Midjourney v6.1, Flux1.1Pro has outperformed in most metrics, particularly in command compliance and image coherence.

Black Forest Labs also stated that Flux1Pro will be upgraded to double its speed while maintaining the same quality. However, Flux1.1Pro remains three times faster than the updated Flux1Pro, achieving an ideal balance between speed and image quality.

Launch of API

The company has also released the beta version of its BFL API, allowing developers to integrate the Flux model into their applications. Black Forest Labs promises customization options, scalability, and competitive pricing.

https://docs.bfl.ml/

As for pricing, Flux1Dev costs $0.025 per image, Flux1Pro is $0.05, and Flux1.1Pro is $0.04 per image. Additionally, these models can be accessed through platforms such as Together.ai, Replicate, fal.ai, and Freepik.

Black Forest Labs was founded this August by developers from Stability.ai, with headquarters in Freiburg, Germany. Their goal is to develop advanced generative deep learning models for image and video applications. Notably, Flux technology has also been used in Elon Musk's Grok AI image generator, although it has sparked some controversy regarding its political uses.

Key Points:
🌟 Flux1.1Pro is the newly launched image generation model, six times faster than its predecessor with higher image quality.
⚡ Black Forest Labs has released the BFL API, enabling developers to integrate the Flux model into their applications.
💰 The Flux models are reasonably priced, with Flux1.1Pro costing only $0.04 per image.

Shakker AI Launches Advanced Image Generation Model Zeno-1 with a Range of Controls

Shakker AI today announced the launch of its latest advanced image generation model—Zeno-1. This model not only inherits the exceptional aesthetics and artistry of its predecessor FLUX.1 but also achieves a leap in stylistic diversity through an internal recommendation system. The launch of Zeno-1 marks a new era for digital art creation tools. It offers a wide range of control tools that enable users to shape and refine their digital artworks in unprecedented ways. Here are the key features of the Zeno-1 model: Palette control: using

Alibaba's Image Generation Model Qwen2vl-Flux Open Sourced, Supporting Image Fusion and Style Transfer

Recently, Alibaba announced the open sourcing of its latest image generation model, Qwen2vl-Flux. This model not only features a variety of functionalities such as editing, fusion, and mixing, but it can also generate highly similar new images based on user-input images or text. Qwen2vl-Flux provides powerful image transformation capabilities. Users simply need to input an image without any text prompts, and the model can generate multiple similar images based on the original. For example, if a user uploads a portrait photo, the model can generate multiple angles.

Can Diffusion Models 'Learn by Analogy'? Alibaba's IC-LoRA Enhances Image Generation Models with Contextual Memory

Recent research from Alibaba's Tongyi Laboratory indicates that existing text-to-image Diffusion Transformer models already possess the ability to generate multiple images with specific relational contexts. With just a little 'guidance', they can achieve a synthesis of understanding and produce high-quality multi-image collections. Traditional Diffusion models are more like 'rote learning' students, requiring massive datasets to generate high-quality images. However, with the support of IC-LoRA, they become 'high-achieving students', capable of learning through analogy.

AI News