AI Setup & API Keys

Get started with Deckyard’s AI-powered features.

New presentation dialog showing the From content option with AI badge for generating slides from text or files

Overview

Deckyard includes AI capabilities to help you create presentations faster and with better content. Features include:

Deck Generation - Create presentations from prompts
Translation - Translate content to other languages
Analysis - Get feedback and suggestions
Description Generation - Auto-generate metadata
Alt Text - Generate image descriptions for accessibility

Supported AI Providers

Deckyard supports multiple AI providers:

Provider	Models	Best For
OpenAI	GPT-4, GPT-4o	General use, image analysis
Anthropic	Claude 3	Long-form content, analysis
Mistral	Mistral Large	European data residency

API Key Configuration

Configure your AI provider by setting the appropriate environment variable.

OpenAI

OPENAI_API_KEY=sk-your-api-key-here

Get an API key from platform.openai.com.

Anthropic (Claude)

CLAUDE_API_KEY=sk-ant-your-api-key-here

Get an API key from console.anthropic.com.

Mistral

MISTRAL_API_KEY=your-api-key-here

Get an API key from console.mistral.ai.

Multiple Providers

You can configure multiple providers simultaneously. Deckyard will:

Automatically detect available providers
Use the best available provider for each task
Fall back to alternatives if one fails

Feature Availability

Different features require different providers:

Feature	OpenAI	Claude	Mistral
Deck Generation	Yes	Yes	Yes
Translation	Yes	Yes	Yes
Analysis	Yes	Yes	Yes
Description Gen	Yes	Yes	Yes
Alt Text (Images)	Yes	No	No

Note: Alt text generation requires OpenAI as it needs vision capabilities.

Checking Provider Status

The application automatically detects which AI providers are available.

In the UI

When you access AI features:

If no provider is configured, you’ll see a setup prompt
Available providers are shown in the AI settings
Features requiring unavailable providers are disabled

Via API

Check available providers:

GET /api/ai/status

Response:

{
  "providers": {
    "openai": true,
    "claude": false,
    "mistral": true
  },
  "features": {
    "deckGeneration": true,
    "translation": true,
    "altText": true
  }
}

Cost Considerations

AI features incur costs based on your provider’s pricing:

OpenAI - Per-token pricing
Claude - Per-token pricing
Mistral - Per-token pricing

Estimating Costs

Typical usage per feature:

Deck generation - Varies based on length (1K-10K tokens)
Translation - Based on content length
Analysis - ~2K tokens per presentation
Description - ~500 tokens

Controlling Costs

Limit access to AI features by user role
Monitor usage in your provider dashboard
Set spending limits in your provider account

Disabling AI Features

To disable AI features entirely, don’t configure any API keys.

To disable specific features:

# Disable all AI
DISABLE_AI=true

# Disable specific features (if supported)
DISABLE_AI_GENERATION=true
DISABLE_AI_TRANSLATION=true

Security

API Key Security

Store API keys in environment variables, not code
Never expose keys in client-side code
Use provider-level restrictions when available

Data Privacy

When using AI features:

Presentation content is sent to the AI provider
Check your provider’s data retention policies
Consider data residency requirements (Mistral for EU)

Troubleshooting

AI Features Not Appearing

Check that an API key is configured
Verify the key is valid (test in provider dashboard)
Check server logs for configuration errors

Errors During AI Operations

Common issues:

Rate limits - You’ve exceeded provider limits
Invalid key - Key is incorrect or revoked
Quota exceeded - You’ve used your monthly allowance

Check your provider dashboard for specific error details.

Slow Responses

AI operations can take several seconds. If consistently slow:

Check your provider’s status page
Consider using a faster model
Reduce input size for generation