Video

---
name: Video
description: Use when the user wants to create, generate, or produce marketing video with AI generation models, AI avatars, or programmatic video frameworks — demos, explainers, or social clips.
category: Content & Copy
version: 2.0.0
tools: []
---

# Video

Help produce marketing videos using AI generation models, AI avatars, and programmatic video frameworks — product demos, explainers, social clips, and ads.

## Before Starting

Search the Knowledge Base for existing product-marketing context first, and only ask what isn't already covered. Gather: **Goal** — video type (product demo, explainer, testimonial, social clip, ad, tutorial), target platform (YouTube, TikTok/Reels/Shorts, website, ads, sales deck), desired length. **Production approach** — human presenter (AI avatar vs. voiceover vs. screen recording), existing footage/assets, need for generated footage, one-off vs. repeatable template. **Technical context** — available tooling/API access, budget (some tools charge per minute of video).

## Choosing Your Approach

| Approach | Best For | Tools | When to Use |
|----------|----------|-------|-------------|
| Programmatic | Templated, data-driven, batch video | Remotion, Hyperframes | Product updates, personalized videos, recurring content |
| AI Generation | Original footage from text/image prompts | Veo, Runway, Kling, Pika | B-roll, hero shots, creative visuals you can't film |
| AI Avatars | Talking-head presenter without filming | HeyGen, Synthesia | Explainers, tutorials, multilingual content |
| Editing/Repurposing | Cutting long-form into short clips | Descript, Opus Clip, CapCut | Podcast/webinar → social clips |

## Programmatic Video

Build videos with code for repeatable, templated, or data-driven output at scale.

**Hyperframes** (HTML/CSS, Apache 2.0, from HeyGen) composes plain HTML documents as frames into a timeline and renders to MP4 — no framework DSL to learn, which makes it LLM-native: coding agents generate better HTML than React components, and rendering is deterministic (same input always produces identical output). Best for product announcements, changelogs, data-driven reports, and personalized outreach videos.

**Remotion** (React) is a more mature, more powerful open-source framework where React components are frames and props drive content; render locally or via Remotion Lambda (AWS) for scale. Best for complex animations, interactive previews, and large batch rendering. Remotion requires a company license for commercial use, versus Hyperframes' Apache 2.0.

| Factor | Hyperframes | Remotion |
|--------|-------------|----------|
| Agent compatibility | Better (plain HTML) | Good (React) |
| Animation complexity | Basic (CSS transitions) | Advanced (Spring, interpolate) |
| Batch rendering | Local | Lambda (AWS) for scale |
| Learning curve | Minimal | Moderate (React + Remotion API) |
| License | Apache 2.0 | Company license for commercial use |

## AI Video Generation

| Model | Resolution | Max Duration | Best For | Cost |
|-------|-----------|-------------|----------|------|
| Veo 3 (Google) | Up to 1080p (4K varies) | Variable | Highest quality, synced audio | API-based |
| Runway Gen-4 | Up to 4K | ~10 sec/gen | Motion control, temporal consistency | $12-76/mo |
| Kling 3.0 | Up to 1080p | Up to 2 min | Volume production, lowest cost | $0.029/sec |
| Pika | 1080p | Short clips | Fast generation, effects | Per-credit |

Sora (OpenAI) has had limited availability and reliability issues — verify current status before recommending it.

Good prompts specify subject + action + camera + style + mood, e.g. "A close-up shot of hands typing on a laptop keyboard, shallow depth of field, warm office lighting, camera slowly pulls back to reveal a modern workspace, cinematic color grading, 4K." Avoid vague subjects, missing camera movement, missing style direction, and requesting readable text in the video (models struggle with it). Use AI generation for an imagined scene, consistent style across clips, or quick B-roll; use stock footage for recognizable real locations; use programmatic video, not AI generation, for specific products/brands.

## AI Avatars

**HeyGen** — best lip-sync and micro-expressions, 230+ avatars, 140+ languages, and an official MCP server so agents can generate avatar videos directly. Tiers: Free (3 videos/mo, 3 min max), Creator (unlimited, 5 min), Business (unlimited, 20 min). Best for product explainers, feature announcements, personalized sales outreach, multilingual content. Custom avatars: upload a 2-5 min video of yourself to create a digital twin that generates videos from text scripts.

**Synthesia** — full-body avatars with expressive body language and built-in script generation from URLs/docs. Best for corporate training, compliance videos, and enterprise presentations where professional tone matters more than realism.

Use an avatar for recurring content, multilingual versions, or personalized outreach at scale. Don't use one for authentic founder content (film yourself), a product UI walkthrough (screen recording), or creative/artistic video (AI generation).

## Editing & Repurposing Tools

**Descript** — transcript-based editing (edit video by editing text); best for cleaning up interviews, podcasts, webinars. **Opus Clip** — auto-clips long videos and scores virality potential; best for long-form → short-form at scale. **CapCut** — effects, captions, platform-native styling; best for TikTok/Reels polish. **Captions.ai** — auto-captions, eye-contact correction, AI dubbing; best for solo talking-head content.

Repurposing workflow: clean up and polish long-form content in Descript, auto-extract the 5-10 best moments with Opus Clip, add captions/effects/platform styling in CapCut, then distribute to TikTok, Reels, Shorts, and LinkedIn.

## Video Production Workflows

**Product demo:** script the key features and value props, screen-record the product flow, add programmatic overlays (titles, callouts, transitions) via Hyperframes/Remotion, generate AI B-roll for establishing shots, record voiceover or use an AI avatar for narration, export at platform-appropriate specs.

**Explainer video:** script the problem → solution → CTA arc, choose a presenter (AI avatar via HeyGen, or voiceover + visuals), build visuals (programmatic slides, screen recordings, AI-generated scenes), always add captions, export landscape for YouTube/website and vertical for social.

**Batch social clips:** build a master template in Hyperframes/Remotion, feed it data (product features, testimonials, stats), render the batch (one template, many variations), add platform-specific captions via CapCut or Captions.ai, then schedule across platforms.

## Agent-Native Video Pipeline

The most powerful setup chains tools an agent can control directly: write the script from product context, generate a templated video via Hyperframes (HTML → MP4) and/or an avatar video via the HeyGen MCP server and/or B-roll via the Veo/Runway API, then assemble the final cut. This works because Hyperframes is plain HTML any coding agent can generate, the HeyGen MCP server lets agents call it directly, and video-model APIs are standard HTTP requests — no manual editing step required.

## Common Mistakes

1. Starting with tools, not strategy — decide what video you need before picking tools
2. AI-generated text in video — models can't reliably render readable text; use programmatic overlays instead
3. Uncanny-valley avatars — invest in a higher avatar tier if quality matters
4. No captions — most social video is watched without sound
5. Wrong aspect ratio — 9:16 for social, 16:9 for YouTube/website, 1:1 for feeds
6. Over-producing — authentic often outperforms polished, especially on TikTok

## Task-Specific Questions

1. What type of video do you need? (Demo, explainer, social clip, ad, tutorial)
2. Do you need a human presenter, or can it be voiceover/text?
3. Is this a one-off or a repeatable template?
4. What platform is it for? (Determines aspect ratio and length)
5. What existing assets are available? (Screenshots, footage, scripts)
6. What's the budget for video tools?

## Related Skills

social (video content strategy, hooks, what to post), ad-creative (paid video ad creative and iteration), copywriting (video scripts and messaging), marketing-psychology (hooks and persuasion in video).