AI Content Generation

Overview

AI-generated content is provided as a starting point to support creativity, efficiency, and exploration. Users are responsible for reviewing and validating all AI-assisted materials — the platform does not guarantee the accuracy of AI-generated content, and it should not substitute for expert instruction.

Available AI Tools

As of version 1.2, the platform supports several AI services:

• Gemini — images, video, 360° media, presentations, AI agents
• OpenAI — alternate image generation, 360 images, video, presentations, AI agents
• Stability AI — the fastest 3D model generation (a picture, then Stable Fast 3D)
• Meshy — 3D models and avatar creation
• Tripo 3D — 3D models and avatars
• MasterpieceX — 3D model generation
• Blender (experimental) — 3D models and rigged avatars built by an AI-written Blender script (OpenAI, Claude or Gemini writes the script; needs Blender installed)
• SketchFab — asset library access (see SketchFab Integration)
• Stable Diffusion — offline image generation, no API key required (see Offline Image Generation)

Note for all of them except for offline Stable Diffusion you will need to get an API key and store it. API keys can be set in the "Options" tab.

API key links:

• Gemini: aistudio.google.com/apikey
• OpenAI: platform.openai.com/settings/organization/api-keys
• Stability AI: platform.stability.ai/dashboard/api-keys (new accounts get 25 free credits; Stable Fast 3D is 10 credits per model)
• Meshy: meshy.ai/settings/api
• Tripo 3D: platform.tripo3d.ai/api-keys
• MasterpieceX: developers.masterpiecex.com/apps (create an app to generate a bearer token)
• SketchFab: sketchfab.com/settings/password (API token is listed on this page — see SketchFab Integration)
• Anthropic (Claude, used for AI agents): platform.claude.com/settings/keys
• ElevenLabs (text-to-speech for AI agents): elevenlabs.io/app/developers/api-keys
• SDXL: Local (offline image generation — see Offline Image Generation)

Content Generation Categories

• 3D Models — Choose between Stability AI, Meshy, Tripo, Blender (experimental) and MasterpieceX, using either text prompts or image-to-3D conversion. For Stability AI, Meshy, Tripo and Blender an Options box appears:
  • Speed / Quality (Meshy, Tripo) switches between two sets of model settings; a line below shows the models that will be used.
  • Text prompts (Meshy, Tripo): make a picture first and turn it into 3D, or go straight from text to 3D.
  • Image generator (Gemini, OpenAI, Stability AI, Meshy or Auto) makes that picture. It is shown while the model builds and becomes the thumbnail when the provider has none.
  • Script written by (Blender): the AI that writes the Blender script.
  • The key field shows the key the selected provider needs. Leave it empty to use a key saved in the Options tab or set with setx.
  • Stability AI and Tripo models are turned to face forward. The model settings are in ai3d_settings.py in the E-Learning Lab folder.

• 360 Media — Generate 360 videos via Veo3, or 360 images through Gemini, OpenAI, or offline Stable Diffusion.

• Standard Video — Veo3 enables video generation.

• Images — Choose Gemini Image 4, Nano-Banana, or OpenAI GPT-Image, with mono or stereo/3D options available (see Converting Monoscopic Images to Stereo).

• Avatars — Generate fully rigged avatars from text prompts or photos using Meshy, Tripo (with a Speed / Quality choice; rigs use Mixamo bone names), or Blender (experimental: an AI-written script builds a stylised humanoid, its skeleton and optional idle and walk animations). An elapsed-time clock shows how long each generation has been running.

Additional Features

Upload PDFs, images, and text to generate contextual quizzes, AI Agent discussions, and supplementary slide content within the scenes panel. See Quizzes and AI Intelligent Agent.