Models

Qwen3 Voice Design: a new voice from one sentence of description, and where to run it

Qwen3-TTS: narration in nine preset voices with pronunciation rules, and where to run it

MOSS: a sound effect from one sentence, and where to run it

ACE-Step 1.5: full songs with sung lyrics from a style prompt, and where to run it

Cosmos3 Nano: the NVIDIA world model for physical scenes and longer shots, and where to run it

Wan2.2 TI2V 5B: a 720p video model small enough for one consumer card, and where to run it

LTX-2.5 Audio-to-Video: a talking video from one still and one voice, and where to run it

LTX-2.5: 1080p video with its own soundtrack from one prompt, and where to run it

LatentSync 1.6: new words for a video you already have, and where to run it

MiniMax H3: a video model that follows your reference pictures, and where to run it

Z-Image-Turbo: the most downloaded image generator in the catalog, and where to run it

SANA-Sprint 1.6B: a 1024-pixel image in one to four steps, and where to run it

OmniGen2: composing one picture from several reference photos, and where to run it

GLM-Image: a two-stage generator built for legible text, and where to run it

DeepGen-1.0: five image jobs in five billion parameters, and where to run it

PaddleOCR: Korean text lifted straight off a photograph, and where to run it

GLM-OCR: a 0.9B document reader ranked first on OmniDocBench, and where to run it

SwinIR: the enlargement baseline everyone measures against, and where to run it

Swin2SR: enlarging a photo two or four times over, and where to run it

Speech Recognition (Clips): timed transcripts for a whole list of clips, and where to run it

Speech Recognition (Video): one video in, one transcript out, and where to run it

ControlNet XL Union: six kinds of control image in one model, and where to run it

ControlNet XL Canny: keep the outlines of a photo while changing everything else, and where to run it

YOLO-S (Fashion Item Detector): 46 garment and trim categories, and where to run it

YOLO-S: object detection with no convolutions at all, and where to run it

FLUX.2 Klein 4B: generating and editing from one model on a consumer card, and where to run it

FLUX Schnell: a text-to-image model that finishes in four steps, and where to run it

Gemma 4 31B: frontier reasoning over a quarter-million tokens, and where to run it

Gemma 4 E2B: a phone-sized model that also hears, and where to run it

SeedVR2 7B: the higher-fidelity restoration checkpoint, and where to run it

SeedVR2 3B: restoring damaged video in a single pass, and where to run it

Helios Distilled: the same minute of video, generated faster, and where to run it

Helios Base: minute-long video from a text description, and where to run it

RF-DETR Keypoint: seventeen joints per person from one photo, and where to run it

RF-DETR Seg Medium: pixel outlines instead of rectangles, and where to run it

RF-DETR Medium: fast, accurate boxes over 80 everyday categories, and where to run it

DETR (License Plate Detector): spotting plates in traffic photos, and where to run it

DETR (Face Detector): a one-class fine-tune for finding faces, and where to run it

DETR: transformer-based detection of everyday objects, and where to run it

Qwen-Layered: splitting a flat photo into transparent layers, and where to run it

Qwen-Inpaint: mask one region, edit only that, and where to run it

Qwen-2511: instruction-based photo editing that keeps the face, and where to run Qwen

SAM 3.1 (Video): Segmentation for videos

SAM 3.1 (Scene): current checkpoint in the SAM 3 concept segmentation line

SAM 3: Tell it what to detect and separate.

SAM 2: Accurate Segmentation

DeepSeek-OCR: where to run it, and what it is good at