FLUX 2 - это прекрасно! Полный детальный разбор и демонстрация возможностей модели. Локально в Comfy

Flux 2 dev is a developer-friendly version of Black Forest Labs' open-weight image generation model (Apache 2.0), containing approximately 32 billion parameters and supporting both text-to-image editing and editing with up to 10 reference images. It combines an improved architecture based on a rectified flux transformer and a new VAE for high-fidelity reconstruction. The model is optimized for local inference, APIs, and cloud environments, making it suitable for production pipelines. Advantages: High-quality images with excellent typography, text rendering, and fine detail; Accurate adherence to prompts, including complex compositions, lighting, and styles, without accidental deviations; Support for multi-reference editing in a single model, open-source customization, and resource savings through FP8; Efficiency for e-commerce, branding, and iterative processes, with the ability to quickly run locally. ✨Support the author: https://dalink.to/ayithedeer ________________________________ ✨Timecodes✨ 00:00 - Introduction 01:42 - What is Flux 2 02:54 - Key Features of Flux 2 04:40 - Workflow for working with Flux 2 in ComfyUI 11:53 - About the new ComfyUI interface 12:38 - What models are there, generation times, and which hardware they are compatible with 15:42 - Text2Image Examples 17:18 - Image2Image Examples and 10 References (Okay, 6) 19:54 - Image2Video Generation Time 21:33 - Image Degradation with Multiple Image2Images 23:19 - Prompt Examples in Russian and Non-Obvious Uses of Flux 2 23:40 - Prompting Instructions for Flux 2 26:22 - Cons and Nuances 28:06 - Flux 2 PRO (ComfyUI API) 29:25 - Conclusions 30:48 - Afterword ________________________________ ✨Links✨ ● Project page: https://flux-2.studio/ ● Download the full model: https://huggingface.co/black-forest-l... ● Download the fp8 model, text encoders, and vae: https://huggingface.co/Comfy-Org/flux... ● Download the GGUF model: https://huggingface.co/city96/FLUX.2-... ● Download the workflow: https://drive.google.com/file/d/1xDVJ... ● Prompting guide for Flux 2: https://docs.bfl.ai/guides/prompting_... Helpful videos on this topic: ● Flux 1 dev and ComfyUI installation:    • Как установить Flux.1 dev на свой компьюте...   ● Qwen Image Edit:    • Qwen Image Edit - это можно назвать нейроф...   ● Analysis of common errors, including GGUF node installation:    • Самые популярные ошибки в ComfyUI - почему...   ________________________________ Tags: What is Flux 2? Flux 2 installation. Flux 2 review in Russian. Neural network for image generation. Free neural network. Local neural network. Neural network for image editing. AI Photoshop. Image editing with a neural network. Free image creation with a neural network. Neural network for photos. Neural network for comics. Neural network for anime. comfyui, flux 2, flux.2, black forest labs, ai image generation, ai, flux2, flux 2 dev, ai art, flux, stable diffusion, ai image editing, comfyui workflow, generative ai, nano banana pro, ai workflow, ai image generator, nano banana, flux ai, ai video, ai news, flux 2 pro, gguf, text generation, hugging face models, aitools, open source ai, flux 2.0, vram optimization, flux image to image, image generation,

Flux.2 Klein - a fast, lightweight model for generating and editing images. Review for ComfyUI
▶︎

Flux.2 Klein - a fast, lightweight model for generating and editing images. Review for ComfyUI

Next Scene - how to maintain the character and atmosphere of a shot from scene to scene. Locally ...
▶︎

Next Scene - how to maintain the character and atmosphere of a shot from scene to scene. Locally ...

Why 24/7 AI Agents Are a Waste of Time
▶︎

Why 24/7 AI Agents Are a Waste of Time

Как VK дошёл до... этого?
▶︎

Как VK дошёл до... этого?

Ideogram 4.0 - A model for typography and design. How to use it locally in ComfyUI.
▶︎

Ideogram 4.0 - A model for typography and design. How to use it locally in ComfyUI.

Your Claude in Python: A-to-Z Google ADK Tutorial
▶︎

Your Claude in Python: A-to-Z Google ADK Tutorial

LongCat-video - генерируем длинные видео без деградации качества и цвета локально в ComfyUI
▶︎

LongCat-video - генерируем длинные видео без деградации качества и цвета локально в ComfyUI

Der schlimmste Pharmakonzern Deutschlands
▶︎

Der schlimmste Pharmakonzern Deutschlands

Käfigkampf vorm Weißen Haus I Musk ist Billionär | extra 3 vom 18.06.2026 · Teil 1/2
▶︎

Käfigkampf vorm Weißen Haus I Musk ist Billionär | extra 3 vom 18.06.2026 · Teil 1/2

Ace Step 1.5 – a model for local music generation. Like Suno, but yours. Review in Russian
▶︎

Ace Step 1.5 – a model for local music generation. Like Suno, but yours. Review in Russian

Flux-2-Dev: Обзор, возможности для маркетплейсов и универсальный рабочий процесс - Стив Ласмин
▶︎

Flux-2-Dev: Обзор, возможности для маркетплейсов и универсальный рабочий процесс - Стив Ласмин

Deep Dive into LLMs like ChatGPT
▶︎

Deep Dive into LLMs like ChatGPT

HiDream AI для ComfyUI. Локальная установка + обзор + сравнение с Flux и Stable Diffusion 3.5
▶︎

HiDream AI для ComfyUI. Локальная установка + обзор + сравнение с Flux и Stable Diffusion 3.5

LTX 2.3 Director: It’s Like Wan 2.1 SVI but for LTX Models. Is It Good, Bad, or What?
▶︎

LTX 2.3 Director: It’s Like Wan 2.1 SVI but for LTX Models. Is It Good, Bad, or What?

Самое опасное когнитивное искажение [Veritasium]
▶︎

Самое опасное когнитивное искажение [Veritasium]

Hunyuan Video 1.5 – the new king of video generation. Review + generation examples. Locally in Co...
▶︎

Hunyuan Video 1.5 – the new king of video generation. Review + generation examples. Locally in Co...

Flux USO - transfer styles from one image to another. Free, locally in ComfyUI.
▶︎

Flux USO - transfer styles from one image to another. Free, locally in ComfyUI.

Wan 2.2 Stable Video Infiniti PRO - generate infinite video on your computer. Comfy guide
▶︎

Wan 2.2 Stable Video Infiniti PRO - generate infinite video on your computer. Comfy guide

You Can't Spot AI
▶︎

You Can't Spot AI

FlashVSR - a low-resolution video upscaler for ComfyUI. Installation + examples
▶︎

FlashVSR - a low-resolution video upscaler for ComfyUI. Installation + examples