Skip to content
FlixMLFlixML Studioopen-source generative AI image and video tool

Wan 2.2 Text-to-Video

Text-to-video with Wan 2.2 — generate a short clip directly from a prompt, no source image.

Workflow id: wan22_t2v · Text → Video

Calling it

POST /api/video/generate
{
  "workflow": "wan22_t2v",
  "prompt": "..."
}

Params

Every input this workflow takes. Your install is the source of truth:GET /api/workflows/wan22_t2v returns these as they are on your machine.

  • prompt · str · required
    Text prompt: subject, setting, and action. Describe big physical action with strong verbs (walks, jumps, turns). Words like subtle, slowly, or gently produce a near-still clip.
  • negative_prompt · str · default bright colors, overexposed, static, blurred details
  • width · int · default 640
  • height · int · default 640
  • length · int · default 81
    Frame count
  • fps · int · default 16
  • seed · int · optional
    Optional. Leave it out and every run is a new random take. Send a seed only to remake an image exactly: the seed a job used is in its record.
  • steps_high · int · default 2
  • steps_low · int · default 2
  • total_steps · int · default 4
    steps_high + steps_low; passed to KSamplerAdvanced.steps
  • cfg_high · float · default 1
  • cfg_low · float · default 1
  • shift · float · default 5
  • sampler · str · default euler
  • scheduler · str · default simple
  • high_model · str · default wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors
    High-noise UNet model
  • low_model · str · default wan2.2_t2v_low_noise_14B_fp8_scaled.safetensors
    Low-noise UNet model
  • vae · str · default Wan2.1_VAE.safetensors
  • clip · str · default umt5_xxl_fp8_e4m3fn_scaled.safetensors
  • lightx2v_high_lora · str · default wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors
  • lightx2v_low_lora · str · default wan2.2_t2v_lightx2v_4steps_lora_v1.1_low_noise.safetensors
  • lightx2v_high_strength · float · default 1
  • lightx2v_low_strength · float · default 1