Scan a Chess Board and Get Instant Analysis (2026)

You've got a chess position in front of you — a diagram in a book, a screenshot a friend sent, a photo of your club game — and you want to know the best move. Scanning the board turns that image into something an engine can actually analyze. Here's how it works and when it's the right tool.

What "scanning a chess board" actually means#

An engine like Stockfish can't look at a picture. It reads FEN — Forsyth–Edwards Notation, a single line of text that describes exactly where every piece sits, whose move it is, and the castling and en-passant rights. So a board scanner's entire job is translation: image in, FEN out.

It uses computer vision to find the board in the image, identify each square's contents, and assemble the FEN string. Once it has that, it can drop the position onto an analysis board and run an engine. Everything downstream — the evaluation, the best-move arrows — is normal engine analysis; scanning is just how the position gets in.

When board scanning is the right tool#

Scanning shines when the position isn't already on a board a tool can read:

• Book and PDF diagrams — you're working through a chess book and want to explore a position further than the author does.
• Screenshots — a position from an article, a forum post, a video frame, or a chat.
• Photos of a physical board — your over-the-board game, a puzzle in a magazine, or a position on a friend's set.
• Puzzle images — a tactic you found as a picture and want to check.

In all of these, there's no digital board underneath — just pixels. Scanning is what bridges that gap.

How to do it with ChessSolve#

ChessSolve includes image-to-FEN scanning built in. You give it an image of a position, its vision model detects the board and pieces and produces the FEN, and then it runs Stockfish 18 — the latest and strongest release of the engine (Stockfish blog) — on the recognized position to show you the evaluation and best moves.

For the cleanest recognition, use a clear image: straight-on angle, good lighting, all the pieces visible, not too blurry. The same things that make a position easy for you to read make it easy for the vision model.

When you don't need to scan at all#

Here's the part a lot of "scan the board" tools won't tell you: if the game is already on Chess.com or Lichess, scanning is the slow way. There's nothing to capture — the board is already digital.

For that case, a live overlay is faster and smoother. Instead of screenshotting the board and feeding the image to a scanner, the overlay reads the online board directly and draws Stockfish's analysis on top of it, updating as the game moves. No capture step, no image at all. That's the everyday workflow for online play and study; scanning is the tool you reach for when the position lives outside an online board.

Pick the tool that matches where the position is#

It comes down to one question — where is the position right now?

• In an image, book, or physical board? Scan it. Image-to-FEN turns it into something analyzable, then the engine does the rest.
• Already on Chess.com or Lichess? Skip scanning and let a live overlay read the board directly.

ChessSolve does both, so you're covered either way. Once you've got the position analyzed, the real work is understanding the engine's answer — see why Stockfish suggests the move it does to turn that best move into something you'll actually remember.