Color quantization reduces the number of distinct colors in an image while preserving its overall visual appearance. A typical 24‑bit photo contains up to 16.7 million colors; quantization brings that down to a limited palette — often 256 colors (8‑bit) — making images more suitable for web delivery, GIF encoding, and retro pixel‑art styling.

Median Cut treats all pixel colors as points in a 3D RGB cube. It repeatedly splits the largest color cluster along its widest axis at the median value, effectively creating smaller and tighter clusters. After reaching the desired number of clusters, each cluster's average color becomes a palette entry. It's fast, widely used, and yields excellent results for general‑purpose quantization.

An 8‑bit palette is the sweet spot for many use cases: it dramatically shrinks file size for PNGs and GIFs, conforms to legacy display limitations, and creates a charming retro aesthetic. Many icons, logos, and web graphics are perfectly readable with 256 or fewer colors, making quantization an effective compression strategy without sacrificing legibility.

Floyd‑Steinberg dithering distributes the rounding error of each quantized pixel to its unprocessed neighbors. This creates a perceptual illusion of intermediate colors and smooth gradients, even with very limited palettes. Enable dithering when you're using fewer than 128 colors or when your image has subtle gradients; disable it when you prefer flat, crisp color regions or need a cleaner look for text and line art.

A 24‑bit image (also called True Color) assigns 8 bits each to Red, Green, and Blue per pixel — supporting 16,777,216 colors. An 8‑bit image uses a palette of up to 256 colors selected from the full RGB space. While 8‑bit images are far smaller, they may exhibit banding or posterization in detailed photos. Quantization bridges the gap by intelligently choosing the best 256 colors for each specific image.

Absolutely! The GIF format is inherently palette‑based (max 256 colors per frame). Use this quantizer to preview how your image will look with a reduced palette before converting to GIF. Download the quantized PNG, then import it into your GIF editor — the color palette will map directly, ensuring no surprises.

Yes — often significantly. PNG files compress far more efficiently when the image uses a limited palette. A 24‑bit PNG might be 500 KB, while the same image quantized to 256 colors could shrink to 80 KB, with minimal quality loss. The exact savings depend on image complexity and the number of colors you choose.

It depends on your goal: 256 for near‑lossless 8‑bit representation, 64–128 for aggressive compression with good fidelity, 16–32 for a distinct posterized or pixel‑art look, and 2–8 for bold stylized effects. Always preview and adjust — this tool updates in real‑time so you can find the best balance.

All processing happens entirely in your browser using JavaScript and the HTML5 Canvas API. Your image never leaves your device — no uploads, no server‑side processing, complete privacy. You can use this tool offline once the page is loaded.

Input: JPEG, PNG, WebP, GIF, BMP, and TIFF (browser‑dependent). Output: The quantized result is always exported as a PNG file, which preserves the reduced palette losslessly and is widely supported across all platforms and applications.