OpenArt adds FLUX.3 text-to-video generation
By Steven Van ·
The tool brings Black Forest Labs' new model to OpenArt, generating 20-second HD clips with native audio and lip-synced dialogue in over a dozen languages.
OpenArt just wired Black Forest Labs' first video model into its creation suite. OpenArt announced on August 6 that FLUX.3 Text to Video is live on the platform, which means you can type a prompt and get back an expressive video clip with native audio, multilingual dialogue, and sound effects that actually respond to what is happening on screen. For a platform best known as an AI image playground, it is a meaningful step into full audiovisual generation.
Black Forest Labs finally has a video model, and OpenArt grabbed it fast
FLUX.3 is a big deal on its own. Black Forest Labs, the German lab behind the FLUX image models that power a large chunk of the AI art world, released FLUX.3 in late July 2026 as its first multimodal model spanning image, video, and audio. The video capability went generally available through the BFL API and select partners on August 4, and BFL claims human raters preferred it over rival models in both text-to-video and image-to-video comparisons, including ByteDance's Seedance line. OpenArt picking it up within days puts the model in front of everyday creators without an API key or a line of code.
The model generates clips up to 20 seconds long in a single pass, at HD resolution with a Full HD upscaling option. That 20-second ceiling matters: most text-to-video tools still top out around 5 to 10 seconds, which forces creators to stitch fragments together and pray the character looks the same in every cut.
The audio is generated with the video, not glued on after
The headline feature is native audio. FLUX.3 renders sound in the same generation pass as the picture, so a door slams on the exact frame it closes and footsteps match the surface a character walks on. That is what OpenArt means by "sounds that respond to the scene," and it is a different experience from generating a silent clip and layering stock foley over it in an editor.
Dialogue is part of the package too. The model produces lip-synced speech in more than a dozen languages, including English, Spanish, French, German, Japanese, Chinese, Portuguese, Hindi, and Italian. For creators making content for international audiences, prompting a character to speak in the viewer's language, with matching mouth movement, collapses what used to be a dubbing workflow into a single text box.
What this unlocks inside OpenArt's suite
OpenArt has been assembling a one-stop creative suite: image generation across dozens of models, custom model training, character consistency tools, and a growing video section. FLUX.3 Text to Video slots into that video suite at openart.ai/suite/create-video/flux-3-video, alongside the platform's other video models.
The practical workflow is simple: describe the shot, the action, and any dialogue or sound you want, then generate. Because OpenArt already handles character and style consistency on the image side, it is well positioned to pair those assets with a video model that understands multi-shot storytelling. Short ads, story beats for social, talking characters, and atmospheric b-roll with built-in ambience are all realistic first projects.
How to try it
FLUX.3 Text to Video is available now on OpenArt's create-video page. OpenArt runs on a freemium credit system, so you can test the model before committing to a paid plan. If you have been waiting for a text-to-video tool where the sound design is part of the generation instead of a second job, this is one of the first places you can use BFL's take on it without touching an API.
