Skip to content

Multi-Modal Content Generation

Multi-Modal Content Generation

Auteryn is the only autonomous agent platform with built-in image and video generation. Create marketing assets, product demos, and social media content as part of your agent workflows.

Unique Capability: Auteryn is the only platform that combines autonomous execution with multi-modal content generation. Competitors like Kimi, Manus, and Devin have no content generation capabilities.


What Can You Generate?

🎨 Images

Create images from text descriptions using Gemini 3 model. Perfect for presentations, diagrams, marketing assets, and web content.

🎬 Videos

Generate videos from text prompts using Veo 3.1 model (premium only). Ideal for product demos, social media, and marketing campaigns.

📊 Presentations

Create interactive presentations with 12 layouts, 10 element types, custom themes, and web publishing. Export as PPTX or publish online.

📈 Charts & Visualizations

Interactive charts with Recharts. Bar, line, pie, area, radar, and scatter plots for data visualization.


Image Generation

Powered by Gemini 3

Auteryn uses Google’s Gemini 3 model for image generation:

  • High quality - Photorealistic and artistic styles
  • Fast generation - 5-10 seconds per image
  • Flexible prompts - Natural language descriptions
  • Multiple aspect ratios - Square, portrait, landscape

Use Cases

Marketing Assets

  • Social media graphics
  • Blog post headers
  • Email campaign images
  • Ad creatives
  • Product mockups
agent.run("""
Create a hero image for our new product launch:
- Modern tech aesthetic
- Blue and purple gradient
- Abstract geometric shapes
- Professional and clean
""")

How It Works

# Simple image generation
image = agent.generate_image(
prompt="Modern office workspace with laptop and coffee",
aspect_ratio="16:9"
)
# Image is automatically saved to sandbox
# /home/agent/output/generated_image_123.png
# Get presigned URL for download
url = agent.export_file(image.path)
# https://cdn.auteryn.ai/...

Rate Limits

PlanImages/RunImages/HourImages/Month
Free510100
Pro10301,000
Business10305,000
EnterpriseCustomCustomCustom

Pricing

  • Credit cost: 50 credits per image
  • Generation time: 5-10 seconds
  • Storage: Included in plan storage

Video Generation (Premium)

Powered by Veo 3.1 (Preview)

Auteryn uses Google’s Veo 3.1 model for video generation:

  • High quality - 1080p resolution
  • Flexible duration - 5-10 seconds per video
  • Natural motion - Realistic physics and movement
  • Multiple styles - Cinematic, documentary, animated

Use Cases

Product Demos

Show your product in action. Rotating views, feature highlights, and usage scenarios.

Social Media

Create engaging content for Instagram, TikTok, and YouTube. Short-form videos that capture attention.

Marketing Campaigns

Generate video ads, explainer videos, and promotional content at scale.

Training Materials

Create instructional videos, tutorials, and educational content automatically.

How It Works

# Generate video (Business/Enterprise only)
video = agent.generate_video(
prompt="Smartphone rotating on white pedestal, studio lighting, 360 degree view",
duration=8, # seconds
aspect_ratio="16:9"
)
# Video is saved to sandbox
# /home/agent/output/generated_video_456.mp4
# Get presigned URL
url = agent.export_file(video.path)

Rate Limits (Strict)

PlanVideos/RunVideos/DayVideos/Month
Free
Pro
Business2550
EnterpriseCustomCustomCustom

Pricing

  • Credit cost: 500 credits per video
  • Generation time: 60-120 seconds
  • Storage: Included in plan storage

Presentation Builder

AI-First Presentation Creation

Auteryn includes a powerful Presentation Builder that lets AI agents create, edit, and redesign interactive slide presentations. Unlike traditional tools, the agent is the primary creator—building complete decks with rich elements while humans get lightweight refinement tools.

agent.run("""
Create a pitch deck for our AI platform:
- Cover slide with hero image
- Problem/solution slides
- Product demo with screenshots
- Traction slide with KPI cards
- Competition comparison
- Timeline roadmap
- Team slide
- Closing with call-to-action
Use modern tech theme with blue/purple colors
""")

Key Features

12 Smart Layouts

Title, content, two-column, image layouts, quote, stat-card, timeline, comparison, and more.

10 Element Types

Headings, text, bullets, images, code, charts, tables, diagrams, shapes, and embeds.

Theme Customization

6 preset themes + custom colors and 8 curated font pairs. Per-slide theme overrides supported.

Interactive Embeds

Embed YouTube videos, Loom recordings, Figma prototypes, and Airtable bases directly in slides.

Hi-Fi Export

Pixel-perfect PPTX export via sandbox rendering. Matches UI exactly with Chart.js and Mermaid support.

Web Publishing

Publish presentations as public web pages with optional password protection. Share via unguessable URLs.

Layouts

The Presentation Builder provides 12 specialized layouts:

Standard Layouts:

  • Title - Cover slides with hero images
  • Content - Standard body slides with vertical flow
  • Two-Column - Split content 50/50
  • Image-Left/Right - Image + text combinations
  • Image-Full - Full-bleed images with overlay text
  • Section-Header - Divider slides
  • Blank - Fully custom positioning

Specialty Layouts:

  • Quote - Pull-quotes with attribution
  • Stat-Card - Up to 4 KPI cards with values and deltas
  • Timeline - Up to 7 milestones across columns
  • Comparison - Side-by-side comparisons (pros/cons, before/after)

Element Types

Text Elements:

  • Headings, body text, bullet lists

Visual Elements:

  • Images (AI-generated or uploaded)
  • Charts (bar, line, pie, area via Recharts)
  • Diagrams (Mermaid flowcharts and diagrams)
  • Shapes (rectangles, circles, arrows)
  • Tables (data tables with formatting)

Interactive Elements:

  • Code blocks with syntax highlighting
  • Embeds (YouTube, Loom, Figma, Airtable)

Themes

6 Preset Themes:

  • Light - Clean white background
  • Dark - Modern dark mode
  • Corporate - Professional blue (default)
  • Creative - Vibrant and bold
  • Minimal - Understated elegance
  • Bold - High contrast

Custom Theming:

  • Override primary, accent, surface, and text colors
  • Choose from 8 curated font pairs
  • Apply per-slide theme overrides for dramatic effects

Font Pairs:

  1. Inter / Inter (default)
  2. Playfair Display / Source Sans Pro (editorial)
  3. Space Grotesk / Inter (modern tech)
  4. Merriweather / Open Sans (traditional)
  5. Oswald / Lato (bold/condensed)
  6. DM Serif Display / DM Sans (elegant)
  7. JetBrains Mono / Inter (technical)
  8. Poppins / Poppins (friendly)

Agent Tools

The agent has access to 5 presentation tools:

1. create_presentation

# Create a complete deck
agent.create_presentation(
title="Q4 Business Review",
slides=[...],
theme={"id": "corporate", "primaryColor": "#0ea5e9"},
aspect_ratio="16:9"
)

2. edit_presentation_slide

# Update a specific slide
agent.edit_presentation_slide(
artifact_id="pres_123",
slide_index=2,
updates={"title": "Updated Title", "add_elements": [...]}
)

3. add_presentation_slide

# Insert a new slide
agent.add_presentation_slide(
artifact_id="pres_123",
slide={...},
position=3 # Insert after slide 2
)

4. remove_presentation_slide

# Delete a slide
agent.remove_presentation_slide(
artifact_id="pres_123",
slide_index=5
)

5. redesign_presentation

# Change theme or apply per-slide overrides
agent.redesign_presentation(
artifact_id="pres_123",
theme={"id": "dark", "primaryColor": "#6366f1"}
)

Export Options

Client-Side Export (Fast):

  • PPTX via pptxgenjs (1-2 seconds)
  • PDF via jspdf + html2canvas
  • Editable text in PPTX
  • Good for quick exports

Hi-Fi Export (Pixel-Perfect):

  • Renders via Playwright in sandbox
  • Screenshots each slide at 1920×1080
  • Assembles PPTX with python-pptx
  • Matches UI exactly (5-15 seconds)
  • Includes Chart.js, Mermaid, embeds
  • Consumes sandbox minutes

Web Publishing

Publish presentations as public web pages:

Features:

  • Unguessable URLs (144-bit entropy tokens)
  • Optional password protection (bcrypt)
  • No login required for viewers
  • Responsive viewer with keyboard navigation
  • Slides served from CDN (fast loading)

Publishing Flow:

# Agent can't publish directly, but users can via UI
# 1. Create presentation with agent
# 2. Click "Publish" in UI
# 3. Optional: Set password
# 4. Get shareable URL: https://presentations.auteryn.ai/{token}

Management:

  • List all published presentations
  • Unpublish to remove public access
  • Republishing creates new token

Playbooks

4 presentation playbooks guide agents:

1. Pitch Deck (presentation-pitch-deck)

  • 10-12 slides
  • Problem, solution, traction, competition, team
  • Uses stat-card for metrics, comparison for competition

2. Quarterly Report (presentation-quarterly-report)

  • 8-10 slides
  • Executive summary, KPIs, highlights, roadmap
  • Heavy use of charts and stat-cards

3. Product Launch (presentation-product-launch)

  • 8-10 slides
  • Product overview, features, pricing, availability
  • Timeline for rollout, embed for demo videos

4. Training (presentation-training)

  • 10-20 slides
  • Learning objectives, content, exercises, resources
  • Comparison for concepts, embed for video tutorials

Use Cases

Pitch Decks

Investor presentations with traction metrics, competition analysis, and team slides.

Business Reviews

Quarterly reports with KPI cards, charts, and executive summaries.

Product Launches

Launch presentations with timelines, pricing comparisons, and demo embeds.

Training Materials

Educational content with video embeds, code examples, and interactive elements.

Sales Presentations

Client proposals with case studies, pricing, and testimonials.

Reports & Analysis

Data-driven presentations with charts, tables, and visualizations.

Example: Complete Pitch Deck

agent.run("""
Create a pitch deck for Auteryn, our AI agent platform:
1. Title slide
- Company name: Auteryn
- Tagline: "AI Agents That Work Like Humans"
- Generate hero image (futuristic AI agents)
2. Problem slide
- 3 pain points businesses face with automation
- Use bullets
3. Solution slide
- How Auteryn solves these problems
- 2-column layout with features
4. Product Demo
- Screenshots of key features
- Image-left layout
5. Traction (stat-card layout)
- 128 customers (+42% QoQ)
- $4.2M ARR (+68% YoY)
- 95% retention
- 4.8/5 rating
6. Competition (comparison layout)
- Left: Auteryn (70+ tools, playbooks, sandbox)
- Right: Competitors (10-20 tools, scripted, browser-only)
7. Roadmap (timeline layout)
- Q2: Hi-fi export
- Q3: Real-time collaboration
- Q4: Enterprise SSO
8. Team slide
- Founder photos and bios
- 2-column layout
9. Quote slide
- Customer testimonial
- "Auteryn replaced two BDRs and a junior dev on day one."
- Attribution: Director of Ops, TechCo
10. Closing slide
- Call to action
- Contact information
Use corporate theme with blue (#0ea5e9) primary color.
Use Playfair Display for headings.
""")

Limits

LimitValue
Max slides per presentation50
Max elements per slide20
Aspect ratios16:9, 4:3
Max published presentations per org200

Pricing

Presentation Creation:

  • Included in all plans
  • No per-presentation charge
  • Uses standard agent execution credits

Hi-Fi Export:

  • Consumes sandbox minutes
  • ~1-2 minutes per 10 slides
  • Charged at browser automation rate

Web Publishing:

  • 100 credits per publish operation
  • Storage included in plan
  • Bandwidth included (CDN-cached)

Chart & Data Visualization

Interactive Charts

Create interactive charts with Recharts:

# Create bar chart
agent.create_chart(
type="bar",
title="Q4 Sales by Region",
data=[
{"region": "North America", "sales": 125000},
{"region": "Europe", "sales": 98000},
{"region": "Asia", "sales": 156000}
],
x_axis="region",
y_axis="sales"
)

Supported Chart Types

  • Bar charts - Compare quantities
  • Line charts - Show trends over time
  • Pie charts - Show proportions
  • Area charts - Cumulative data
  • Scatter plots - Correlations
  • Radar charts - Multi-dimensional data

Features

  • Interactive - Hover tooltips, zoom, pan
  • Responsive - Adapts to screen size
  • Exportable - Download as PNG or SVG
  • Customizable - Colors, labels, legends

Integration with Workflows

Automated Marketing Campaign

# Agent workflow combining multiple content types
agent.run("""
Create a complete marketing campaign for our new product:
1. Generate hero image (16:9, modern tech aesthetic)
2. Create product demo video (10 seconds, feature highlights)
3. Generate 3 social media images (1:1, different angles)
4. Create presentation (10 slides, product overview):
- Title slide with hero image
- Problem/solution slides
- Feature highlights with stat-cards
- Pricing comparison
- Customer testimonials (quote layout)
- Call to action
5. Export all assets to /output/campaign/
Use consistent blue/purple brand colors throughout.
""")

Automated Report Generation

# Weekly report with charts and visuals
agent.run("""
Generate weekly performance report:
1. Fetch data from database
2. Create 5 charts (revenue, users, engagement, retention, growth)
3. Generate summary image with key metrics
4. Create presentation with Presentation Builder:
- Title slide
- Executive summary
- KPI stat-cards (revenue, users, growth)
- Individual chart slides
- Trends and insights
- Action items
5. Export as hi-fi PPTX
6. Email to team@company.com
Run every Monday at 9 AM.
""")

Best Practices

Image Generation

Do:

  • Be specific in prompts (“modern office” vs “office”)
  • Specify style (“photorealistic”, “illustration”, “minimalist”)
  • Include colors and mood
  • Mention composition (“centered”, “close-up”, “wide angle”)

Don’t:

  • Use vague prompts
  • Request copyrighted characters or logos
  • Generate inappropriate content
  • Exceed rate limits

Video Generation

Do:

  • Keep videos short (5-10 seconds)
  • Describe motion clearly (“rotating”, “zooming”, “panning”)
  • Specify lighting and environment
  • Test prompts before bulk generation

Don’t:

  • Request complex scenes (keep it simple)
  • Generate videos with people (privacy concerns)
  • Exceed daily limits
  • Use for real-time applications (generation takes time)

Presentations

Do:

  • Use templates for consistency
  • Include data sources
  • Keep slides focused (one idea per slide)
  • Use charts instead of tables when possible

Don’t:

  • Overcrowd slides with text
  • Use too many different fonts/colors
  • Skip title and conclusion slides
  • Forget to cite data sources

Competitive Advantage

Auteryn vs. Competitors

FeatureAuterynKimi K2.5Manus AIDevin AI
Image Generation✅ Gemini 3
Video Generation✅ Veo 3.1
Presentation Builder✅ 12 layouts, themes, web publish
Chart Generation
Integrated Workflows

Only Auteryn combines autonomous execution with multi-modal content generation. Create complete marketing campaigns, reports, and presentations—all automated.


API Reference

Image Generation

# Generate image
image = agent.generate_image(
prompt="Description of image",
aspect_ratio="16:9", # "1:1", "16:9", "9:16", "4:3"
style="photorealistic" # "photorealistic", "illustration", "minimalist"
)
# Returns: ImageResult
# - path: /home/agent/output/image_123.png
# - url: presigned URL (24h expiry)
# - size: file size in bytes

Video Generation

# Generate video (Business/Enterprise only)
video = agent.generate_video(
prompt="Description of video",
duration=8, # seconds (5-10)
aspect_ratio="16:9" # "1:1", "16:9", "9:16"
)
# Returns: VideoResult
# - path: /home/agent/output/video_456.mp4
# - url: presigned URL (24h expiry)
# - duration: actual duration
# - size: file size in bytes

Presentation Builder

# Create presentation
presentation = agent.create_presentation(
title="Presentation Title",
slides=[
{
"layout": "title",
"title": "Main Title",
"elements": [
{"type": "heading", "content": "Subtitle"},
{"type": "image", "content": "https://..."}
]
},
{
"layout": "stat-card",
"title": "Key Metrics",
"elements": [
{"type": "heading", "content": "$4.2M"},
{"type": "text", "content": "Revenue"},
{"type": "text", "content": "+68% YoY"}
]
},
{
"layout": "content",
"title": "Slide Title",
"elements": [
{"type": "text", "content": "Body text"},
{"type": "bullets", "content": ["Point 1", "Point 2"]}
],
"notes": "Speaker notes for this slide"
}
],
theme={"id": "corporate", "primaryColor": "#0ea5e9"},
aspect_ratio="16:9"
)
# Edit a slide
agent.edit_presentation_slide(
artifact_id=presentation.id,
slide_index=1,
updates={
"title": "Updated Title",
"add_elements": [{"type": "text", "content": "New content"}]
}
)
# Add a slide
agent.add_presentation_slide(
artifact_id=presentation.id,
slide={"layout": "content", "title": "New Slide", "elements": [...]},
position=2 # Insert after slide 1
)
# Remove a slide
agent.remove_presentation_slide(
artifact_id=presentation.id,
slide_index=3
)
# Redesign theme
agent.redesign_presentation(
artifact_id=presentation.id,
theme={"id": "dark", "accentColor": "#10b981"}
)
# Returns: PresentationArtifact
# - id: artifact ID
# - title: presentation title
# - slide_count: number of slides
# - theme: applied theme
# - aspect_ratio: "16:9" or "4:3"

Troubleshooting

Image Generation Failed

Symptoms: “Image generation failed” error

Solutions:

  1. Check prompt for inappropriate content
  2. Verify rate limits not exceeded
  3. Simplify prompt if too complex
  4. Check credit balance

Video Generation Timeout

Symptoms: Video generation takes >2 minutes

Solutions:

  1. Simplify prompt (fewer elements)
  2. Reduce duration (try 5 seconds)
  3. Check server status
  4. Retry during off-peak hours

Low Quality Output

Symptoms: Generated content doesn’t match expectations

Solutions:

  1. Be more specific in prompts
  2. Add style keywords (“professional”, “modern”)
  3. Specify colors and composition
  4. Try multiple variations

Resources


Questions?

  • How long does image generation take? 5-10 seconds per image.
  • Can I generate videos on Pro plan? No, video generation requires Business or Enterprise plan.
  • What image formats are supported? PNG (default), JPEG, WebP.
  • Can I use generated content commercially? Yes, you own all generated content.
  • Are there content restrictions? Yes, no inappropriate, copyrighted, or harmful content.
  • How many slides can a presentation have? Maximum 50 slides per presentation.
  • Can I publish presentations publicly? Yes, with optional password protection via unguessable URLs.
  • What’s the difference between client and hi-fi export? Client export is fast (1-2s) but approximate. Hi-fi export is pixel-perfect (5-15s) and matches the UI exactly.
  • Can I embed videos in presentations? Yes, YouTube, Loom, Figma, and Airtable embeds are supported.

View all FAQs →