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

# Generation Help & Tips

> Write better prompts, pick the right node, and fix results you don't like.

export const DemoVideo = ({src, poster, caption, aspectRatio = "16 / 10", className = ""}) => {
  return <figure className={`not-prose my-6 w-full ${className}`}>
      <div className="overflow-hidden rounded-2xl border border-gray-950/10 bg-gray-100 dark:border-white/10 dark:bg-white/5" style={{
    aspectRatio
  }}>
        <video src={src} poster={poster} loop autoPlay muted playsInline className="h-full w-full object-cover" />
      </div>
      {caption && <figcaption className="mt-2 text-center text-sm text-gray-500 dark:text-gray-400">
          {caption}
        </figcaption>}
    </figure>;
};

A clear prompt and the right node make the biggest difference to your results. Here's how to get great output on the canvas — and what to do when a generation misses or stalls.

<DemoVideo src="/videos/help/generating.mp4" poster="/images/posters/help/generating.jpg" caption="Refining a prompt and re-running a node for a better result" />

## Write better prompts

Most results improve the moment your prompt gets more specific:

* Describe the **subject** clearly — who or what is in the scene.
* Name a **style** — photo-real, illustration, cinematic, flat, hand-drawn.
* Set the **mood** — lighting, color, energy, time of day.
* Connect a **reference image** when you want a consistent look across generations.
* **Iterate** — run it, read the result, and change one detail per attempt so you can see what each tweak does.

## Pick the right node

Each node type is tuned for a different kind of output, so matching the task to the node saves time and credits:

* [Image](/canvas/nodes/image) for stills, frames, and concept art.
* [Video](/canvas/nodes/video) for motion and short clips.
* [Audio](/canvas/nodes/audio) for speech, music, and sound effects.
* [Vector](/canvas/nodes/vector) for logos and scalable graphics.
* [Storyboard](/canvas/nodes/storyboard) for multi-shot sequences.
* [Flam](/canvas/nodes/flam) for interactive, motion-rich HTML pieces.

Browse the full lineup in the [nodes overview](/canvas/nodes/overview), or let the [AI assistant](/canvas/ai/overview) build the node graph for you.

## Fix results you don't like

* **Anchor the look** — reuse the same reference image and keep your style wording steady between runs.
* **Change one variable** — adjust a single detail per attempt instead of rewriting the whole prompt.
* **Touch up instead of re-rolling** — for images, the [image tools](/canvas/image-tools/overview) let you inpaint, outpaint, upscale, or remove the background without regenerating from scratch.

## When a generation fails or gets stuck

A failed generation is not a lost credit — Flowy refunds it automatically, so you can try again.

<Steps>
  <Step title="Give it time">
    Video takes noticeably longer than images. Let it finish before assuming it's stuck.
  </Step>

  <Step title="Retry the node">
    Run it again. Many failures are temporary and clear on a second attempt.
  </Step>

  <Step title="Simplify your prompt">
    Trim extra detail and keep the core idea. A shorter, clearer prompt often succeeds where a crowded one stalls.
  </Step>

  <Step title="Check your credits">
    Generations need available credits to start. Check your balance in [Settings → Billing](/billing/credits).
  </Step>
</Steps>

<Note>
  If failures keep happening, work through [troubleshooting](/help/troubleshooting). To confirm a credit came back after a failure, see [refunds](/help/refunds).
</Note>