FLUX 3 Video Generation Model Launches with Up to 20 Seconds of Native 1080p Video

ITHome reported on August 6 that Black Forest Labs issued an announcement yesterday (August 5), officially launching the FLUX 3 video generation model. It can generate videos up to 20 seconds long at a native 1080p resolution, with performance superior to models such as ByteDance’s Seedance 2.0 and Minimax’s H3. ITHome previously reported that the FLUX 3 model entered early access on July 23, using a unified architecture to jointly learn images, video, and audio. FLUX 3 can output videos up to 20 seconds long in a single generation, complete with native audio. According to the company, the model supports text-to-video, image-to-video, video-to-video, continuation of input videos and audio, keyframe-to-video, multilingual dialogue, and multi-shot sequencing.

According to data published in the official blog post, the model scored 1,135 points for text-to-video generation, surpassing the Gemini Omni Flash and Minimax H3 models. For image-to-video generation, it scored 1,051 points, surpassing Seedance 2.0 and Minimax H3.

In terms of pricing, usage is calculated based on the number of seconds in the generated video. Draft mode (fast/economical) is limited to HD resolution. Text-to-video or image-to-video generation costs $0.06 per second (ITHome note: approximately RMB 0.41 at the current exchange rate), while video-to-video generation costs $0.12 per second (approximately RMB 0.81 at the current exchange rate).

Under normal HD conditions, text-to-video or image-to-video generation costs $0.17 per second (approximately RMB 1.1 at the current exchange rate), while video-to-video generation costs $0.41 per second (approximately RMB 2.8 at the current exchange rate).

In Full HD mode, text-to-video or image-to-video generation costs $0.29 per second (approximately RMB 2 at the current exchange rate), while video-to-video generation costs $0.53 per second (approximately RMB 3.6 at the current exchange rate).