Skip to main content

​
Seance Mode

Seance Mode is the core experience of SeanceAI, allowing you to have deep, personal conversations with any of our 60+ historical figures. Each conversation is powered by advanced AI that brings these legendary minds back to life with authentic personalities, beliefs, and speaking styles.

​
How It Works

Seance Mode creates an immersive experience where you can:
1

Select a Historical Figure

Browse our collection of 60+ figures spanning Ancient World to Modern Era. Use search and era filters to find the perfect conversation partner.
2

Start Your Conversation

The figure “emerges from the cosmic void” with a personalized greeting. Choose from starter questions or ask anything you’d like.
3

Engage in Dialogue

Each figure responds with their authentic personality, vocabulary, and perspectives from their historical era. They’ll reference their real experiences, works, and documented beliefs.
4

Get Contextual Suggestions

After each response, receive AI-generated follow-up questions tailored to your conversation to keep the dialogue flowing naturally.

​
Key Features

​
Authentic Historical Personas

Each figure is carefully crafted with:
  • Personality traits: From Einstein’s playful wit to Cleopatra’s regal charisma
  • Speaking style: Vocabulary and expressions appropriate to their era and background
  • Historical knowledge: Awareness only of events up to their death year
  • Documented beliefs: Their real philosophical, political, and personal values
When you ask about modern concepts they couldn’t have known, figures react with genuine curiosity appropriate to their time period.

​
Streaming Responses

Experience real-time conversation with smooth streaming text that appears as the AI generates it, just like a real chat. API Endpoint: POST /api/chat/stream 
const response = await fetch('/api/chat/stream', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({
    figure_id: 'einstein',
    message: 'What inspired your theory of relativity?',
    history: conversationHistory,
    model: 'google/gemma-3-12b-it:free'
  })
});

​
Conversation History

SeanceAI maintains context throughout your conversation:
  • History tracking: Up to 20 previous messages kept in context
  • Auto-save: Conversations automatically saved to browser storage
  • Resume anytime: Return to previous conversations from the saved conversations panel
  • Branch conversations: Create alternate timelines from any message point
History is limited to the last 20 messages to optimize AI performance and response quality.

​
AI Model Selection

Choose from multiple AI models organized by capability:
Fast, responsive models perfect for most conversations:
  • Gemma 3 12B
  • Gemma 3 27B
  • Llama 3.3 70B
  • Llama 3.1 405B
API Endpoint: GET /api/models 
{
  "models": [
    {
      "id": "google/gemma-3-12b-it:free",
      "name": "Gemma 3 12B",
      "tier": "swift"
    }
  ],
  "default": "google/gemma-3-12b-it:free"
}

​
Conversation Controls

  • Back button: Return to figure selection
  • Theme toggle: Switch between dark and light modes
  • Model selector: Change AI model mid-conversation

​
Sharing & Export

Generate a shareable link or use the Web Share API to share your conversation. Links include the full conversation history and can be opened by anyone.
// Share via Web Share API (mobile-friendly)
await navigator.share({
  title: `Conversation with ${figureName}`,
  text: conversationText,
  url: window.location.origin
});
Copy the entire conversation as formatted text for pasting into documents or notes.
Download your conversation as a .txt file with proper formatting and timestamps.

​
Branch Management

Create alternate conversation paths from any message:
  1. Click the branch button (🌿) next to any message
  2. Name your new branch
  3. Continue the conversation in a different direction
  4. Switch between branches using the branch selector
Branches are perfect for exploring “what if” scenarios or comparing different conversation directions.

​
Example Conversation Flow

// 1. Select figure
const figure = await fetch('/api/figures/einstein').then(r => r.json());

// 2. Send message
const response = await fetch('/api/chat', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({
    figure_id: 'einstein',
    message: 'How did you feel when you learned your work contributed to the atomic bomb?',
    history: [],
    model: 'google/gemma-3-12b-it:free'
  })
});

const data = await response.json();
console.log(data.response); // Einstein's authentic response

// 3. Get follow-up suggestions
const suggestions = await fetch('/api/suggestions', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({
    figure_id: 'einstein',
    history: [{ role: 'user', content: '...' }, { role: 'assistant', content: data.response }],
    last_response: data.response
  })
}).then(r => r.json());

console.log(suggestions.suggestions); // ["Tell me more...", "What was your perspective?", ...]

​
Tips for Better Conversations

  1. Ask about their experiences: Reference their real historical events and works
  2. Explore their beliefs: Question their documented philosophical and political views
  3. Challenge their ideas: Create engaging debates and philosophical discussions
  4. Use starter questions: They’re crafted to elicit interesting responses
  5. Follow suggestions: AI-generated follow-ups help maintain conversation flow

​
Technical Details

​
Rate Limiting

SeanceAI includes intelligent rate limit handling:
  • Automatic retries: Up to 3 retries with exponential backoff (2s, 5s, 10s)
  • Model fallback: Automatically tries alternative models if primary is rate-limited
  • Graceful degradation: Falls back from streaming to regular API if needed

​
System Requirements

  • Modern web browser with JavaScript enabled
  • Internet connection for API requests
  • localStorage support for conversation saving
  • Optional: Web Share API support for mobile sharing

​
Next Steps

Historical Figures

Browse all 60+ available figures by era

Dinner Party Mode

Host conversations with multiple figures

Conversation History

Learn about saving and managing conversations

Build docs developers (and LLMs) love

Get started for freeTalk to us