Prompting Guides
GPT-5.2 Prompting Guide
Latest model best practices: structured outputs, tool use, chain-of-thought, system prompt design.
Codex Prompting Guide
Optimizing prompts for the Codex CLI agent: code generation, debugging, refactoring, multi-file edits.
Meta Prompting Guide
Using AI to improve your prompts: automated prompt optimization, self-critique patterns, prompt chaining.
Realtime Prompting Guide
Best practices for the Realtime API: voice interactions, streaming responses, latency optimization.
GPT-4.1 Prompting Guide
Instruction following, long context handling, agentic coding patterns.
Code Examples
The cookbook includes runnable Jupyter notebooks for:
- Function calling and tool use
- Embeddings and semantic search
- Fine-tuning workflows
- Batch processing
- Vision and multimodal
- Text-to-speech and speech-to-text