AI Foundations
Your starting point. Tour the Skytells platform, understand both APIs, make your first Prediction and Inference calls, and handle errors in production.
Who this is for
Developers who are new to the Skytells platform — or to AI generation APIs in general. No prior AI experience required.
What you'll be able to build after this path
By the end of these six modules, you'll be able to:
- Navigate the Skytells platform and obtain an API key
- Choose the right model and estimate costs before writing any code
- Understand the difference between the Prediction API and Inference API, their request/response schemas, and when to use each
- Make your first image/video/audio generation call using the Prediction API
- Make your first LLM chat, streaming, stateful conversation, and embedding call using the Inference API
- Handle errors from both APIs with production-grade retry logic, webhooks, and security practices
Your learning path
Module 1 — Platform Tour
Console, Dashboard, and API overview. Get your API key and explore the platform without writing code.
Module 2 — Models & Billing
Explore the full model catalog — TrueFusion, partner models, video, audio. Understand per-prediction billing and estimate costs.
Module 3 — The Two APIs: Schemas & Concepts
Learn the Prediction API and Inference API side by side. Understand request/response schemas, the type field, and how to tell responses apart.
Module 4 — Your First Prediction
Make your first real Prediction API call in cURL, Python, and TypeScript. Poll for completion and read output URLs.
Module 5 — Your First Inference Call
Call the Inference API using the OpenAI-compatible interface. Stream tokens, hold stateful conversations, and generate embeddings.
Module 6 — Error Handling & Best Practices
Implement retry logic, set up webhooks for async predictions, secure API keys, download outputs safely, and pass the production readiness checklist.
Start with Module 1 →