#cost optimization
3 total
Guides
12 minOpenAI's Agents Optimization cookbook: a four-round playbook for cheaper support agents
A new official cookbook tutorial on agent optimization: measure a baseline, change one thing per round, and accept savings only after quality holds. Four optimization rounds and a nine-step tuning order, item by item.
Read articlePrompt
7 minPrompt Engineering for Reasoning Models: When to Use Them, How to Prompt, How to Save Tokens
Reasoning models need different prompts than regular GPT models: the official best practices -- simple and direct beats technique stacking, no step-by-step coaching, seven task types worth delegating, and reasoning-item token savings.
Read articleGPT Models
6 minreasoning.effort in Practice: Making Reasoning Depth a Tunable Parameter
Same model, adjustable thinking depth. This tutorial covers the reasoning.effort syntax, per-tier scenarios, how it interacts with max_output_tokens and the incomplete status, and the cost consequences of getting it wrong.
Read article