> ## Documentation Index
> Fetch the complete documentation index at: https://docs.zevor.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Getting started

> From zero to your first video in Zevor: create an account, get credits, pick a mode, describe and download. We recommend starting from a concrete recipe.

<Steps>
  <Step title="Create an account">
    Sign up at [zevor.ai](https://zevor.ai) with email or Google.
  </Step>

  <Step title="Get credits">
    Generating uses credits. Free accounts start with none, so pick a
    plan at [zevor.ai/pricing](https://zevor.ai/pricing) to start
    generating. See [Credits and plans](/en/credits-and-plans).
  </Step>

  <Step title="Pick a mode">
    In **Create**, choose the mode based on what you have: just an
    idea, a photo, a motion clip, etc. Each mode has its own page.
  </Step>

  <Step title="Describe and generate">
    Write a clear prompt, set quality and format, and hit **Generate**.
    The result shows in the panel; past generations stay in your
    gallery.
  </Step>

  <Step title="Download and post">
    Download the MP4 or image and upload to TikTok, Reels or Shorts.
  </Step>
</Steps>

## Your first recommended flow

If it's your first time and you don't know where to start, follow a
full recipe from the common ones:

<CardGroup cols={2}>
  <Card title="Build an AI influencer" icon="user" href="/en/recipes/ai-influencer">
    Generate your AI character and turn it into vertical videos.
  </Card>

  <Card title="Recreate a viral clip with Motion" icon="person-running" href="/en/recipes/motion-clip">
    Film a movement and apply it to a photo in under 10 minutes.
  </Card>

  <Card title="Restore an old photo" icon="wand-sparkles" href="/en/recipes/restore-photo">
    Repair and colorize an old photo in 2 steps.
  </Card>

  <Card title="Talking photo with your audio" icon="microphone" href="/en/recipes/talking-photo">
    A photo + your voice: video with synced lips.
  </Card>
</CardGroup>

## Prompt tips

* Be specific about scene, lighting and framing.
* One detail per sentence: the AI follows short sentences better.
* If a result almost works, reuse it (Re-run) and tweak the prompt
  instead of starting over.

<Note>
  Prompt quality matters more than the model. Before switching modes
  or quality, try rewriting the description.
</Note>