AI Video Generator

Four ways into video

Each tool has its own feature page and a matching studio app. Open any app without signing in. Subscribe when you are ready to spend credits on a generation.

• Text to Video

  Write a prompt and get a short clip. Wan 2.7 also accepts optional driving audio.

  Wan 2.2 · Wan 2.7

• Image to Video

  Animate a still frame. Wan 2.7 supports end frames, continuation, and audio.

  Wan 2.2 · Wan 2.7

• Audio to Video

  Let a soundtrack drive the motion. Optional first-frame image or prompt.

  LTX 2.3

• Video to Video

  Retarget motion onto a character, or relip existing footage to new audio.

  Kling Motion Control · VEED Lipsync

Which tool should you open?

Start with text-to-video when you only have a description. Use image-to-video when you already like a frame and want motion. Audio-to-video is for music or dialogue-led clips. Video-to-video is for motion control or lipsync on footage you already shot. If you are still choosing between stills and motion, generate a keyframe in text-to-image first, then animate it. That path usually wastes fewer credits than rewriting a long video prompt from scratch.

Habits that save credits

Keep the first video prompts short: one subject, one camera move, one action. Watch the full clip before you download. Soft source images make soft video, so upscale a weak still before image-to-video when pixels are the problem. For lipsync and motion control, clean driver footage beats fancy settings.

Credits and access

Studio pages are public to browse. Generating spends credits from a Starter, Creator, or Pro plan. Sign in before checkout so the subscription and credit balance attach to your account. Video jobs usually cost more than stills, so treat early runs as tests with a clear goal.