Search the full model catalog.

Undistilled 4B base model of FLUX.2 [klein] for text-to-image and multi-reference editing. Apache 2.0 licensed and built for fine-tuning and research, with more varied outputs than the distilled models.

FLUX.2 Klein 4B is a 4 billion parameter rectified flow transformer from Black Forest Labs that generates images from text and edits with up to 4 reference images. Open weights under Apache 2.0.

Black Forest Labs' 9B FLUX.2 [klein] model, step-distilled to 4 inference steps, that generates images from text and edits them with up to 4 reference images in a single checkpoint.

Animates a still image into a five-second clip with the distilled LTX-2 checkpoint from Lightricks, generating the picture and its soundtrack together, including dialogue written into the prompt.

Turns one photo of a person into a dance video. Pick one of 35 trend routines, such as Mente Ma, Supernova or Six Seven, and the person in the photo performs it.

Kling 2.5 Turbo turns a text prompt, optionally with a start image, into a silent 5 or 10 second video, tuned for highly dynamic motion and close prompt adherence. Pro mode adds a last-frame target.

The lighter Turbo variant of FLUX.2 [dev], Black Forest Labs' 32B open-weight image model. It works from a prompt alone or edits and combines up to 5 reference images.

Transform images into animated videos using the Wan2.2 Animate Replace model. Generate dynamic video content from static reference images and driving videos.

Animates a still character image with the body motion and facial expressions of a driving video. The character keeps its own look and background, and the clip runs as long as the driving video.

Turns a word or two of text into a short 3D title video. The letters become neon tubes on a wet street, gold foil balloons, cake decorations or one of 19 other scenes.

Turns a product photo and a short caption into an animated promo video, with the text set in the scene, such as a neon sign, 3D letters or a holiday card. There are 41 presets, including Black Friday and Christmas sets.

Turns a product photo and a logo into a short branded video ad. Pick one of 12 scene presets, from a street billboard or golden storefront to a logo drawn in cappuccino foam.

Tokenizer-free TTS from OpenBMB for English and Chinese. It infers prosody from the meaning of the text and clones a voice zero-shot from a short clip and its transcript.

Meituan's 6B-parameter text-to-image model for Chinese and English prompts. Quoted words go through character-level encoding, so headlines and signs come out legible.

Puts the person from your photo in a vertical scene with a music player card showing a YouTube track's title, then plays that track under the image as a short video.

Z-Image Turbo is the distilled, lighter variant of Tongyi-MAI's 6B-parameter Z-Image family. It turns a text prompt into photorealistic images and can write English and Chinese text inside them.

Pruna VACE is Pruna AI's optimized build of VACE, the all-in-one video model from Alibaba's Tongyi Lab. It makes clips of up to 81 frames at 480p or 720p, with up to 3 reference images to keep characters consistent.

Qwen Image Edit Plus is the Qwen team's multi-image editing model, offered here by Pruna. It edits one photo from a written instruction or merges people and objects from two photos into one scene.

Wan Image Small is a smaller text-to-image model that Pruna offers for prototyping and large batches. It returns one still per prompt and adds a 21:9 widescreen frame to the usual aspect ratios.

P-Image is Pruna's own text-to-image model. Pruna documents it for exact prompt adherence and for writing requested words into the picture. Each request returns one image in one of seven aspect ratios.