Prompt weighting assigns different importance levels to different parts of a text prompt. Syntax examples: AUTOMATIC1111 uses (word:weight), Midjourney uses ::weight suffix, ComfyUI supports various notations. How it works: Multiply token embeddings by weight before cross-attention. Higher weight = stronger influence on generation. Use cases: Emphasize key subjects ((cat:1.4) sitting on couch), de-emphasize elements ((background:0.7)), balance competing concepts. Weight ranges: 1.0 is default, 0.5-1.5 typical range, extreme weights (>2.0) can cause artifacts. Nested weights: ((word)) often equals (word:1.1) squared, syntax varies by tool. BREAK keyword: Some tools use BREAK to separate prompt sections into different conditioning chunks. AND operator: Combine multiple prompts with equal influence. Per-word vs per-phrase: Can weight individual tokens or entire phrases ("detailed landscape:1.3"). Trade-offs: Heavy weighting can distort generations, reduce coherence. Best practices: Use subtle weights (0.8-1.2), test iteratively, fix prompt issues directly. Useful for fine-tuning composition and emphasis.