Gemini

1. Models

   Gemini model families, capabilities, versions, and choosing between Pro, Flash, and Flash-Lite.

2. API keys and authentication

   Creating and securing API keys and using them with the Gemini API and GenAI SDK.

3. Text generation

   generateContent, system instructions, generation config, multi-turn chat, and streaming responses.

4. Streaming

   Server-streamed incremental responses and handling partial output.

5. Structured outputs

   Constraining responses to JSON schemas for predictable parsing.

6. Function calling

   Declaring tools, letting the model request function calls, and returning results.

7. Thinking

   Thinking budgets and levels, thought summaries, and thought signatures for reasoning models.

8. Long context

   Million-token context windows, token counting, and long-context usage patterns.

9. Image, document, video, and audio understanding

   Passing images, PDFs, video, and audio as inputs, and the Files API for large media.

10. Built-in tools and grounding

    Google Search grounding, code execution, URL context, and file search tools.

11. Embeddings

    Gemini embedding models, task types, and output dimensionality for retrieval and semantic search.

12. Live API

    Low-latency bidirectional voice and video sessions over WebSockets.

13. Context caching

    Implicit and explicit caching to cut cost and latency for repeated large prompts.

14. Batch API

    Asynchronous high-volume requests at reduced cost.

15. Rate limits, pricing, and optimization

    Cost and latency levers including flex and priority inference, quotas, and usage tiers.

Each section links to the official page. Titles and summaries by Extra Credit; the docs belong to their authors.