Chess.com vs Lichess vs Chessable for improving: which one, for what

For improving, use Lichess to play and analyze (free, unlimited engine), Chess.com if your friends and tournaments are there, and Chessable only when you already understand an opening and want to drill its moves. None of the three explains why you lost; that gap is filled by a coach, human or AI, and by your own analysis.

The question "Chess.com vs Lichess vs Chessable" assumes three competitors. They aren't. Lichess is where you play, Chess.com is where your friends are, and Chessable is where you memorise. Each is excellent at its job and mediocre at the other two. The mistake most improvers make is expecting one of them to do the job none of them does: tell you why a move was bad in words you can use in your next game.

Disclosure first: I build LearnChess, which competes with pieces of all three. It gets its own section at the end, where it fits a job, and is not ranked against the three. Every price and feature below was checked on 4 November 2026 on each vendor's own site or its US App Store listing; where I could not confirm something, the text says so.

The one-table answer

Checked 4 November 2026. The Chess.com App Store listing shows several prices per tier without labelling the billing period, so the monthly figures are ranges; web and regional prices differ. Lichess's Patron page says: "Lichess is entirely free, forever, and for everyone."

Lichess: best for playing and analyzing

Lichess is the only one of the three that is free in the strict sense. No ads, no paid tier, open source, run by a non-profit on donations. Patron status buys a badge on your profile and nothing else. Everything an improver uses after a game is included: free server-side Stockfish analysis (40 games a day, far more than anyone uses), with the evaluation graph, the blunder list and a "learn from your mistakes" mode that makes you find the better move yourself.

Beyond analysis it has an opening explorer with master games and Lichess games by rating band, studies (shared annotated boards that work as free opening files), puzzle themes with a dashboard of the ones you fail most, Puzzle Storm for speed, and endgame practice against the engine. For a player under 1800 that is the complete toolkit, at zero.

The playing pool is smaller than Chess.com's. This matters only at the edges: very high ratings, odd time zones, unusual time controls. Under 2000, at 15+10, you wait seconds. Ratings run higher on Lichess than on Chess.com for the same player, so never compare numbers across sites.

The weakness is guidance. Lichess will tell you that 14.g4 dropped the evaluation by two pawns and show the engine's line. It will not say that g4 opened the king, that the bishop on f5 needed trading first, or what the same mistake looks like in your other games. Perfect data, no interpretation. Many players run the analysis after every game, nod at the red moves, and learn nothing, because reading an engine line is a skill nobody taught them. I plateaued around 1600 on Lichess for a long time, and reading those lines was the part I could not do.

Chess.com: best for community and content

Chess.com is where the chess world is: the biggest pool, the clubs, the streamers, the titled-player events, the friends who sent you the invite. If the people you want to play are there, play there; a playing site you don't open does nothing for your chess.

It also has the most content of the three: thousands of video and interactive lessons, Puzzle Rush and Puzzle Battle, bots at every level, and Game Review, which annotates your game with move labels (Brilliant, Great, Mistake, Blunder), an accuracy percentage and, on Diamond, written coach explanations for every move. The mobile apps are the best-made in chess.

The weakness is the paywall ladder. On the check date, Chess.com's support pages gave free members "1 full game review analysis per day" and a daily puzzle cap. Gold removes the puzzle and lesson caps. Platinum gives "unlimited game review and analysis". Diamond adds the coach explanations and the Insights dashboard. So the two things an improver needs most, reviewing every game and seeing trends across games, sit on the two most expensive tiers. The membership page promised to "Increase Your Rating 2.7X Faster With Premium!", a claim nobody can check; read it as marketing.

A second, quieter weakness is accuracy theatre. The move labels and the accuracy percentage feel like feedback and are easy to chase. An 85% game you lost on move 40 and a 70% game you lost to a one-move blunder are different problems with different fixes, and the number does not tell you which. The Diamond coach explanations help, but they are templates triggered by engine features ("this hangs your knight", "this weakens your king") and rarely reach the level of plan or structure. Reliable and shallow.

If you already pay for Diamond, use it fully: Insights for the trends, Game Review for every serious game. If you don't, Lichess does the analysis part for free, and importing a Chess.com game there takes a minute.

Is Chessable worth it? Best for drilling an opening you already understand

Chessable is a course marketplace with a move trainer attached. You buy or open a course, read the author's text, and MoveTrainer quizzes you on the moves, repeating each one at growing intervals until you play it from memory. The library is the largest in chess, over a thousand courses, with authors up to the world elite. For the specific job of turning a line you understand into a line you play on autopilot, nothing else is as good. (Chessable belongs to Chess.com; its announcement says free Chessable courses are coming inside Chess.com, for Diamond members after the beta. I could not confirm the current state of that rollout.)

The cost has two parts. PRO is $11.99 a month or $74.99 a year in the US App Store, with a 7-day trial, and mostly buys training conveniences: scheduling control, variation previews, statistics. The courses are the real spend. Free "Short & Sweet" versions exist by the hundreds and are a fair taste, but a full repertoire from a known author typically costs tens of dollars, the biggest around a hundred, and both colours means several of them. Chessable runs frequent sales; the sticker price is rarely what people pay.

Worth it depends on one question: do you already understand the opening you want to drill? If yes, Chessable is the best tool for the next step, and the price is fair for a complete, maintained file. If no, it is the most expensive way to acquire a memorised repertoire that collapses on move 9 when your opponent leaves the course. The understanding lives in the course text, and the training screen quietly encourages you to skip it: the arrows are right there, the text is a tap away, and the streak counter rewards moves, not reasons. Most people drill the arrows.

The test for understanding, as opposed to memory: cover the board after move 8 and say, in words, what each side wants to do next. If you can't, more drilling will not help; the fix is a guide or course text that explains the plan. Our opening guides are written for that first step, and the method is in how to build an opening repertoire you actually understand.

What none of them does: tell you why

Put the three side by side and the gap is obvious. Lichess gives you the engine's verdict with no words. Chess.com gives you words, on Diamond, generated from templates. Chessable gives you an author's words about the opening, written before you ever played a game, so they are never about your mistake.

What an improver needs after a loss is a sentence like: "You traded the dark-squared bishop on move 12, which made the hole on f6 permanent; after that the knight on e5 could never be challenged, and the g4 push on move 19 was just the moment it became fatal." That sentence connects a decision to a structure to a result. A human coach produces it in the first five minutes of a lesson. None of the three sites does, and the three together don't either.

Between roughly 1000 and 1800, most adult players already have more data than they use: the engine (Lichess, free), the move labels (Chess.com), sometimes a drilled repertoire (Chessable). What they lack is the step from "this move was -2.0" to "I do this kind of thing, and here is the pattern." That step is interpretation, and no site built for playing, community or drilling has a reason to build it.

So the honest answer to "which one for improving" is: pick a playing site by where your games are, run analysis where it is free, drill openings only after you understand them, and get the why from somewhere else: a coach, a study group, a club member 300 points stronger, an annotated game collection, or an AI coach grounded in the engine rather than guessing.

Chess.com vs Lichess vs Chessable for improving: which is best?

None is best overall. By goal:

"I want a rating and a community." Chess.com to play, because the pool and the friends are there. Export your games to Lichess for free analysis, or pay Platinum to stay in one place. Skip Chessable unless a specific course is on your list. Cost: $0 to about $10 a month.

"I want to understand my games." Lichess to play and analyze; slow games, 15+10 or longer, because blitz builds reflexes, not understanding. After each loss, before you open the engine, write one sentence about where it went wrong, then check it against the evaluation graph. The gap between your sentence and the engine's verdict is your curriculum. Add a source of explanations: a coach, annotated games in your openings, or an AI coach that reads the actual board. Cost: Lichess is free; the explanation layer is where you spend.

"I want an opening repertoire." Read first, drill second. Read a guide that explains the plan behind each move until you can say what both sides want at move 8. Then drill the moves in Chessable, or a free trainer, until they are automatic. Then play the opening in rapid games for a month and lose with it; every loss in a structure you are learning shows you a plan you did not know yet. Cost: a course or two, plus PRO only if you drill daily.

"I'm a beginner." Lichess or Chess.com, whichever your friends use, and nothing else for the first few hundred games. Both teach the rules, both have puzzles and bots. Chessable is a waste of money before you know which openings you like. Buy nothing until you have a problem you can name.

The pattern across all four: two tools, not three. One place to play, where your games will actually happen, and one tool that turns the games into something you understand. The third subscription is where people start collecting apps instead of improving. The longer version, across ten apps, is the best chess apps in 2026, sorted by job.

Where LearnChess sits

LearnChess is mine, so weigh this paragraph accordingly. It is not a playing site; it imports your games from Lichess and Chess.com. It sits in the gap the three leave: the AI coach answers questions about the position on the board in plain language, grounded in the live Stockfish evaluation, master-game databases and opening theory rather than a template or a guess; the analysis board is a simple branching board with the engine built in, free for every account; the opening guides explain the plan behind each move; and the repertoire tool makes you prove a line (play it from memory, answer engine-checked questions about the ideas) before it counts, then re-tests it after 90 days. I built it because, stuck around 1600 on Lichess, I wanted something that would decipher the engine lines and let me ask "why not the knight?" in words. If that is your gap too, the free plan includes the engine and ten coach messages, enough to see whether the explanations are the kind you want. If your gap is a playing pool or a thousand courses, the three above are the tools.