- Full-video cloning
- Face, lines, B-roll, captions & effects
- Batch scale
- 10, 50, or 100+ per run
- Open-source control
- Fork, modify, self-host
- Agent compatibility
- Claude, Codex, any agent
- Model freedom
- BYOK or Hypit API
- Software cost
- $0 open-source core
Clone any viral video
with AI agents.
1 command, 100 variants, 100M views.
Trusted by teams from
Not just the script. The whole video.
Your agent can edit the footage, dialogue, B-roll, captions, and effects—all anchored to words instead of seconds
Hover to see the full source
1
<script>2
3
<segment id="manifest-rule">4
<take>{manifest-rule.semantic.take}</take>5
</segment>6
7
<segment id="real-estate-rule">8
<take>{real-estate-rule.semantic.take}</take>9
</segment>10
11
<segment id="bitcoin-rule">12
<take>{bitcoin-rule.semantic.take}</take>13
</segment>14
15
</script>16
17
<broll-track>
</broll-track>18
<caption-track document={story.caption}/>19
<emoji-track during=program>
</emoji-track>20
<media-track>
<item during={story.selection.lifestyle} />
</media-track>Hover to see the full source
1
<script>2
3
<segment id="opening-question">4
<take>{opening-question.semantic.take}</take>5
</segment>6
7
<segment id="daily-creatine">8
<take>{daily-creatine.semantic.take}</take>9
</segment>10
11
<segment id="arms-are-asking">12
<take>{arms-are-asking.semantic.take}</take>13
</segment>14
15
</script>16
17
<caption-track document={story.caption}/>18
<media-track>
<item during={story.selection.lifestyle} />
</media-track>Hover to see the full source
1
<script>2
3
<segment id="ronaldo">4
<take>{ronaldo.semantic.take}</take>5
</segment>6
7
<segment id="messi">8
<take>{messi.semantic.take}</take>9
</segment>10
11
</script>12
13
<caption-track document={story.caption}/>14
<broll-track>
<item during={ronaldo.hair-gel} />
</broll-track>Hypit vs others
| Feature |  |  |  | |
|---|---|---|---|---|
| Full-video cloning | Face, lines, B-roll, captions & effects | Remix inside Arcads | Ad templates, 15s max | Generates clips only |
| Batch scale | 10, 50, or 100+ per run | Limited by paid plans | Limited by credits | Batch it yourself |
| Open-source control | Fork, modify, self-host | Closed source | Closed source | Closed model and API |
| Agent compatibility | Claude, Codex, any agent | Arcads Agent + MCP | MCP, no public API | API only |
| Model freedom | BYOK or Hypit API | Platform catalog | Platform catalog | Seedance only |
| Software cost | $0 open-source core | Reported ~$110+/mo | $19–$129/mo | Pay per second |
- Full-video cloning
- Remix inside Arcads
- Batch scale
- Limited by paid plans
- Open-source control
- Closed source
- Agent compatibility
- Arcads Agent + MCP
- Model freedom
- Platform catalog
- Software cost
- Reported ~$110+/mo
- Full-video cloning
- Ad templates, 15s max
- Batch scale
- Limited by credits
- Open-source control
- Closed source
- Agent compatibility
- MCP, no public API
- Model freedom
- Platform catalog
- Software cost
- $19–$129/mo
- Full-video cloning
- Generates clips only
- Batch scale
- Batch it yourself
- Open-source control
- Closed model and API
- Agent compatibility
- API only
- Model freedom
- Seedance only
- Software cost
- Pay per second
Start with anything.
Start with a reference video, an idea, or a proven template.
01 / Clone
Clone a video format
Drop in any reference video. Hypit rebuilds its shots, pacing, captions, and effects with your content.
Clone a video
02 / Agent
Create with an agent
Describe the video you want. Your agent plans, generates, and assembles it.
Start with an agent
03 / Templates
Use a proven template
Start with a podcast, street interview, dialogue, video call, or any proven format.
Browse them on GitHub