Descriptive text added to images that screen readers and search engines use to understand image content. Good alt text describes content and purpose, not just its appearance — “Team celebrating product launch” is better than “people in office.” Framer lets you add alt text directly in the image properties panel. See How to add Alt Tags to images.

A compressed image format best suited for photographs and complex images with many colors and gradients. JPEG compression is lossy, meaning some quality is sacrificed for smaller files. Use JPEG for photos but prefer PNG for graphics with sharp edges, text, or transparency.

A lossless image format supporting transparency, best for graphics, logos, and images with sharp edges or text. PNG files are larger than JPEG for photos but preserve quality perfectly through editing and compression. Use PNG for graphics with transparency or when image quality is paramount.

Temporary content indicating where final content will appear, helping visualize layouts before content is ready. Placeholders can be lorem ipsum text, gray boxes, or sample images. Replace placeholders with real content before launch—they can accidentally go live.

High-resolution screens with twice or more the standard pixel density, displaying sharper text and images. Retina displays require higher resolution images—typically 2x the displayed size—to appear crisp. Framer automatically serves appropriate image sizes for different display densities.

Largest Contentful Paint (LCP) measures loading performance by tracking when the largest visible text or image element appears on screen.

A Context Window is the maximum amount of tokens a model can process at once, including instructions, conversation history, and retrieved data.

Retrieval-Augmented Generation (RAG) enriches model outputs by fetching external knowledge at runtime and conditioning generation on it.

Grounding is the practice of constraining generation with verifiable sources so outputs are accurate, attributable, and context-specific.

Multimodal AI combines understanding and generation across different modalities, enabling richer interfaces and cross-media reasoning.

Prompt Enhancement is an AI-assisted rewrite step that expands or clarifies user intent before generation. It often works with a Prompt Template and explicit System Prompt constraints.

Style Reference lets you guide the aesthetic of generated assets by pointing the model to example visuals. It is frequently combined with Reference Image inputs in Text-to-Image Generation workflows.

A Reference Image is a conditioning input that guides composition, structure, or aesthetics during generation. It is central to Style Reference workflows and Multi-image Conditioning.

Multi-image Conditioning uses several images as control inputs for one generation task, improving consistency across outputs. It extends single Reference Image workflows in Text-to-Image Generation.

Generative Expand increases image boundaries and predicts plausible continuation beyond original edges. It is commonly used alongside Generative Fill in broader Text-to-Image Generation workflows.

A Diffusion Model creates images through iterative denoising steps conditioned on prompts and controls. It is the backbone of many Text-to-Image Generation systems and can be steered by Classifier-Free Guidance (CFG).

ControlNet augments diffusion generation with explicit structural conditions such as edges, depth, or pose. It improves control in Diffusion Model systems and often uses cues from Image Segmentation.

Segment Anything Model (SAM) produces masks from points, boxes, or text-like prompts for rapid object selection. It underpins modern Image Segmentation workflows and improves control in Reference Image editing.

Image Segmentation partitions an image into labeled regions to isolate objects or areas for editing. It is core to Segment Anything Model (SAM) workflows and precision operations like Generative Fill.

Prompt-to-Prompt Editing changes specific image attributes by adjusting textual instructions while preserving overall scene structure. It is closely related to Prompt Enhancement and iterative Text-to-Image Generation.

InstructPix2Pix applies natural-language editing commands to existing images while retaining layout context. It extends ideas from Prompt-to-Prompt Editing within practical Text-to-Image Generation pipelines.

DreamBooth is a personalization method that fine-tunes a model to generate consistent renderings of a chosen subject. It is a specialized form of Fine-tuning used in Subject-Driven Generation.

Textual Inversion introduces new concept tokens by learning embeddings that map to visual ideas. It is lightweight compared to full training and connects closely with Embeddings and DreamBooth workflows.

Subject-Driven Generation aims to keep a specific person, product, or character consistent across new generated scenes. It is often implemented with DreamBooth and guided by a Reference Image.

The ability of AI models to translate intent—expressed as prose, pseudocode, or examples—into executable source code. Code generation underpins vibe coding, autocomplete tools, and agentic workflows. Generated code should always be reviewed for correctness, security, and maintainability, since models can produce plausible-looking but incorrect implementations.