ChessGrader / Guides / Best free review tools
The best free chess game review tools, tested in 2026
We checked six ways to analyze a chess game for free: ChessGrader, Chess.com, Chesssigma, Chess It Up, WintrChess, and Lichess. Here is what each one gives you, where the limits sit, and which job it does best.
There's no single best free review tool. There's a best tool for volume, a best tool for a polished daily report, a best tool for unlimited Chess.com-style grades. This list includes ChessGrader — we make it — so every entry gets the same treatment: what it does, a "best for" line, and an honest limitation. Including ours.
How we checked: We tested the public analysis flow and documentation for each tool on August 2, 2026, without starting a trial. Feature limits can change, so this page carries a visible test date. For game-rating output, we also ran the same five public games through ChessGrader and Chesssigma; that benchmark is below. It is a comparison, not a claim that either site's estimate is ground truth.
1. Lichess — the default answer
If someone asks "where do I analyze positions for free," Lichess is still the default answer. Its analysis board runs Stockfish locally in the browser and accepts FEN or PGN. Logged-in players can also request server-generated computer analysis for games, subject to Lichess's account rate limits. You get an eval graph, accuracy, average centipawn loss, and every Inaccuracy, Mistake, and Blunder flagged. Around it sits the best free study ecosystem in chess: an opening explorer over billions of games, a masters database, and Studies for building shareable annotated analysis. The whole site is open source and donation-funded, so none of this is a funnel to a paid tier.
Best for: volume analysis and serious study. The catch: the report is spartan. No Brilliant, Great, or Miss labels, no estimated game rating, no coach. Lichess tells you where you went wrong and leaves the encouragement to you.
2. ChessGrader — unlimited Chess.com-style grading
ChessGrader is our answer to the gap between the two big sites: Chess.com's report style at Lichess's price. Every move graded with the full vocabulary — Brilliant through Blunder, including Great, Forced, and Miss — plus accuracy, an estimated game rating, an eval graph, and mistake drills. Unlimited, no signup. Stockfish 17.1 runs as WebAssembly in your browser at a fixed 100,000 nodes per move, with critical moments re-verified at 600,000 nodes, so your games never touch a server. Import from Chess.com or Lichess by username, or paste any PGN. The V2-inspired expected-points model, boundaries, and safeguards are published on the methodology page — if you think a grade is wrong, you can inspect the method.
Best for: unlimited Game Review-style reports with published methodology. The catch: analysis speed depends on your device (about a minute per game on a modern phone), it's standard chess only — no variants — and there's no opening explorer or server database.
3. Chess.com free tier — the best coach experience
Chess.com's Game Review is the most polished post-game report in chess, and the free tier includes one per day. You get the famous move labels, an accuracy score, an estimated game rating, and a coach persona walking you through the key moments. If you play on Chess.com, it's one click after every game, and for a single daily game it's hard to beat the experience.
Best for: the friendliest single review of today's game. The catch: one review per day, full coach commentary and blunder details in self-analysis need a paid membership, and although the V2 concepts and ordinary boundaries are public, the fitted grading and CAPS2 accuracy models are private.
4. Chesssigma — free analysis with AI coach features
Chesssigma offers free, unlimited game analysis powered by Stockfish 17, with no signup needed to analyze. It imports games by Chess.com or Lichess username or from a pasted PGN, and produces move classifications from Brilliant to Blunder, accuracy scores, and eval graphs. Its distinguishing bet is the coaching layer: an AI "Supercoach" that reads across your games to build a training plan, plus puzzle cycles and an opening trainer. It's the most feature-dense of the independent tools.
Best for: unlimited analysis plus an ambitious training product in one place. The catch: the move-grading methodology is not published, and the coaching and training product extends beyond the no-signup analysis flow. Like ChessGrader, its Stockfish analysis runs in the browser.
5. Chess It Up — a configurable analysis board
Chess It Up combines a free analysis board with direct Chess.com and Lichess username import and PGN paste. Its free full-game analysis runs Stockfish NNUE at depth 12, grades moves from Brilliant to Blunder, and reports accuracy and average centipawn loss. It also shows best lines and opening-book moves.
Best for: players who want a traditional analysis board around the game report. The catch: ads fund the free tier, and deeper analysis up to depth 25 is a paid feature.
6. WintrChess — the open-source game report
WintrChess grew out of WintrCat's popular "Game Report" project and does one thing: free game reports with move classifications, no signup. Paste a PGN or pick a game from your Chess.com or Lichess account and it flags everything from brilliancies to blunders. The whole project is open source under GPL-3.0, which puts it in rare company with Lichess on the transparency front.
Best for: a quick, no-frills free report from an open-source project. The catch: it's a volunteer-maintained project — at the time of writing the repository is openly seeking maintainers — and it's a single-purpose report tool without drills, databases, or study features.
Side by side
| Lichess | ChessGrader | Chess.com (free) | Chesssigma | Chess It Up | WintrChess | |
|---|---|---|---|---|---|---|
| Unlimited reviews | Local board yes; server reports rate-limited | Yes | No (1/day) | Yes | Yes | Yes |
| Signup required | For your own games | No | Yes | No (for analysis) | No | No |
| Move labels | Inaccuracy/Mistake/Blunder | Brilliant → Blunder | Brilliant → Blunder | Brilliant → Blunder | Brilliant → Blunder | Brilliant → Blunder |
| Accuracy score | Yes | Yes | Yes | Yes | Yes | Yes |
| Estimated game rating | No | Yes | Yes | Yes (estimator) | No published estimator | Yes |
| PGN import | Yes | Yes | Yes, through Analysis | Yes | Yes | Yes |
| Methodology public | Yes | Yes, fully | No | No | No | Open source |
| Engine runs | Browser and servers | Your browser | Their servers | Your browser | Your browser | Mixed |
Our reproducible Chesssigma comparison
ChessGrader includes an estimated single-game rating, so we checked it against the closest comparable output from Chesssigma. The sample contains five public games across novice, club, and expert ratings, producing ten player estimates. We recorded both tools' output on July 20, 2026 and kept the game list in the project's regression corpus.
| Public game | Player ratings | Chesssigma estimates | ChessGrader estimates |
|---|---|---|---|
| ashsko–Sigurd_2000 | 232 / 246 | 550 / 750 | 550 / 750 |
| katius71–battlebott | 759 / 764 | 900 / 1150 | 1200 / 1250 |
| sh722–127chetan | 1792 / 1719 | 2300 / 2400 | 2250 / 2200 |
| abusch38–Icemankids | 2081 / 2060 | 2200 / 2100 | 2100 / 2100 |
| Bolt46–macenri | 1531 / 1528 | 1350 / 1300 | 1500 / 1650 |
Across the ten estimates, ChessGrader's mean absolute difference from Chesssigma was 125 Elo and the largest difference was 350 Elo. That does not make Chesssigma ground truth. Single-game ratings are noisy, and the two products use different private or published models. The useful result is that ChessGrader stays in the same broad scale while remaining anchored to the player's real rating.
One number that never transfers between tools: accuracy. Each site uses its own formula, so an 85% here isn't an 85% there. Compare only within one tool.
How to actually choose
- You analyze many games and study openings: Lichess. Nothing else combines unlimited deep analysis with a real database.
- You want Chess.com-style graded reports for every game you play: ChessGrader. Unlimited, no signup, public methodology, games never uploaded.
- You play one serious game a day on Chess.com: the free daily Game Review is genuinely good. Use it, then use a free tool for the rest.
- You want an AI training plan built from your games: Chesssigma is the one betting on that.
- You want a familiar analysis board with configurable depth: Chess It Up. The free tier is useful; deeper analysis is paid and the page carries ads.
- You want a quick open-source report with zero commitment: WintrChess.
- You have OTB games in a notebook: anything with PGN paste works; ChessGrader's PGN analyzer grades them without an upload.
Frequently asked questions
Do I need to pay for chess game review at all?
No. Lichess offers a free local analysis board and limited server-generated reports, while ChessGrader, Chesssigma, Chess It Up, and WintrChess offer free game reports with Chess.com-style move labels. Paying mainly buys deeper settings, coaching, convenience, or an integrated training product.
What's the best free alternative to Chess.com's Game Review?
For the closest match to the Game Review format, ChessGrader, Chesssigma, Chess It Up, and WintrChess all provide free reports with move labels and accuracy. ChessGrader additionally publishes its expected-points model and safeguards and runs Stockfish 17.1 in your browser. For exploratory analysis and study tools, Lichess remains the standard free answer.
Are free chess analysis tools as accurate as paid ones?
Many free and paid tools use Stockfish, but their results are not automatically identical. Engine version, NNUE network, search budget, hardware, and the formulas that turn evaluations into labels and accuracy scores all matter. A free tool with a deep search, like Lichess at roughly 2 million nodes per move, can analyze more deeply than many paid reviews.
Can I use these tools for over-the-board games?
Yes, if the tool accepts PGN input. Transcribe your OTB game into PGN, then paste it into Lichess, ChessGrader, Chesssigma, Chess It Up, or WintrChess. ChessGrader, Chesssigma, and Chess It Up state that their engine analysis runs in your browser.
Why do the same game and moves get different accuracy scores on different sites?
Each site uses its own search settings and accuracy model. Lichess publishes its method; Chess.com's exact CAPS2 formula is private; ChessGrader publishes a CAPS2-inspired approximation calibrated against public reviewed-game benchmarks. The scores can differ materially and are only comparable within a single tool.
Is it safe to type my username into these tools?
Generally yes. Tools that import by username read the public game archives that Chess.com and Lichess expose for every account. There is no login or password involved, so the tool cannot access or change anything on your account.
