WAN 2.2 T2V Cascade · 14B Parameters

Text to Video
AI Generation Studio

Generate cinematic videos from text using a dual-stage cascade diffusion model. Real-time progress, live logs, complete control.

🔑

★ Prompt Examples ▾

✦ Prompt

0 chars

◈ Video Size & Duration

Width px

Height px

Duration s

Presets:

⚡ Generation Mode

✦ Quality Mode ~30 min

20 steps · CFG 3.5 · Higher quality output

⚡ Turbo Mode ~5 min

4 steps · CFG 1.0 · Lightning LoRA applied

⬡ Models & LoRA ▾

Connect API key to load available models.

High Noise Model Stage 1

Low Noise Model Stage 2

High Noise LoRA

Low Noise LoRA

CLIP / Text Encoder

VAE

CtrlEnter to generate · Turbo: ~5 min

🎬

Connect your API key, write your prompt, and click Generate. First run loads models (~3–5 min), then video generation begins.

WAN 2.2 Cascade 14B Params Dual-Stage SSE Live

Model loading

3–5 min

Stage 1 — High noise

varies

Stage 2 — Low noise

varies

VAE decode

1–2 min

⚠

◈ Job History 0

No completed jobs yet. Generate a video to see it here.