来源：https://www.promptingguide.ai/techniques 爬取日期：2026-03-22

Prompting Techniques

Prompt Engineering helps to effectively design and improve prompts to get better results on different tasks with LLMs.

While the previous basic examples were fun, in this section we cover more advanced prompting engineering techniques that allow us to achieve more complex tasks and improve reliability and performance of LLMs.

Available Techniques

The guide covers the following 18 prompting techniques:

1. Zero-shot Prompting — Prompting without any examples
2. Few-shot Prompting — Providing demonstrations/examples in the prompt
3. Chain-of-Thought Prompting — Enabling reasoning through intermediate steps
4. Meta Prompting — Using prompts to generate or improve prompts
5. Self-Consistency — Sampling multiple reasoning paths and selecting the most consistent answer
6. Generate Knowledge Prompting — Generating relevant knowledge before answering
7. Prompt Chaining — Breaking complex tasks into sequential sub-tasks
8. Tree of Thoughts — Exploring multiple reasoning pathways systematically
9. Retrieval Augmented Generation (RAG) — Combining retrieval with generation
10. Automatic Reasoning and Tool-use (ART) — Automating reasoning with tool integration
11. Automatic Prompt Engineer (APE) — Automated prompt optimization
12. Active-Prompt — Actively selecting the most informative examples
13. Directional Stimulus Prompting — Using hints/cues to guide generation
14. Program-Aided Language Models (PAL) — Offloading computation to code
15. ReAct — Combining reasoning and acting with external tools
16. Reflexion — Self-reflection and iterative improvement
17. Multimodal CoT — Chain-of-thought for multimodal inputs
18. Graph Prompting — Prompting with graph-structured data