Best AI chess coach: ranked picks for your needs

The best AI chess coach depends on the job: start with Cody for live board questions, Chessiro for follow-up explanations of your games, ChessHere for a weekly training focus, or Chessvia for voice-led review [17][19][20][15]. This is a ranked shortlist by use case, not a tested rating-gain leaderboard. Try a real position before paying.

The order prioritises a clear coaching interaction: asking about the board, following up on your own mistakes and turning review into practice. The final section keeps specialist training tools together so an opening course or child-focused playing site is not mistaken for a conversational coach. ChessCore is academy management software and is not a player-app alternative.

Cody is the first pick for a live question about the displayed position. Its official page says it follows the board and accepts text or voice follow-ups [17]. Start here: https://codychess.com/ai-chess-coach. Best for: someone who wants to ask why before moving on.

The free plan includes unlimited text coaching, a 10-minute voice welcome and then 5 voice minutes weekly. Club is USD 14.99 monthly and Master USD 29.99 monthly on the global page [18]. These are displayed web prices, not an India checkout quote.

The documented starting workflow is to place a position on the board, make a move and ask what to notice [17]. For your trial, use a position from a completed game that you still find confusing. Explain what you intended, then ask what the opponent can do against that plan. This is a more revealing test than asking for the best move without any context.

After the answer, change the position and ask the same question again. Check whether the explanation names the pieces and squares now on the board. Board awareness matters because a plausible general lesson is not enough when a tactical detail changes the verdict. If the explanation and variation disagree, keep the line visible and ask for the concrete reply.

A voice interface suits a player who loses their train of thought while typing. It may be less useful in a shared room or on a commute. Test the environment where you will actually study, including whether you can interrupt, replay the idea and make your own notes. Paid voice time is worth considering only if speaking materially improves that session.

Do not make a large library migration your first task here. Cody's coaching page centres on board coaching rather than file imports or a native mobile app. Begin with the board interaction, then confirm any import route, account connection or device support you need before committing to a subscription.

Chessiro suits post-game questions: it reports engine-grounded explanations and position-specific follow-ups, with PGN, Chess.com and Lichess game import [19]. Start here: https://www.chessiro.com/ai-chess-coach. Best for: a player with recent games and a specific mistake to understand.

Engine review is free and unlimited. The weekly free coaching allowance is 3 move explanations, 2 coach chats and 2 Smart Shuffle sessions; paid plans raise those limits [19]. Do not confuse the unlimited engine layer with unlimited conversation.

Choose a completed game and identify where your own confidence diverged from the result. Perhaps you believed the attack was winning, or thought a trade solved your defensive problem. Give that intention to the coach after review. The useful outcome is an explanation of the decision you actually made, not only a list of errors sorted by size.

Follow up by asking what you should have checked before choosing the move. Then close the explanation and reconstruct the reply on the board. If you can repeat the wording but cannot find the move, you have a reading result rather than a usable chess lesson. Bring the unresolved position to your next human lesson instead of asking the same question until a comforting answer appears.

A small free conversation allowance can still be enough to judge the interaction. Spend it on the moment that matters rather than asking broad questions about improvement. Keep a note of which explanations helped and which needed correction. Upgrade only when the allowance interrupts a routine that is already producing useful notes.

For a coach recommending this route, separate technical success from learning success. Importing a game and generating text proves the workflow runs. The student explaining the missed threat in their own words is a different check. Ask for that explanation in the next lesson before expanding the amount of assigned review. Check whether a native mobile app exists before assigning phone-based homework.

ChessHere reports a loop from its own games to debriefs, a weekly focus and daily practice drawn from the player’s mistakes [20]. Start here: https://www.chesshere.com/ai-chess-coach. Best for: someone who wants a defined next task after reviewing a game.

The official page says starter coaching, a daily puzzle, weekly focus and progress report are free, and work on web, iOS and Android [20]. It describes play within ChessHere rather than external PGN or account import. Check the paid coaching limits before upgrading.

This workflow is most relevant when your problem is indecision after analysis. You finish reviewing a loss, recognise several faults, then spend the next session choosing what to study instead of practising. A suggested focus gives you something concrete to accept, modify or discuss with a coach. Treat the focus as a proposal supported by games, not as an unquestionable diagnosis.

Check the evidence behind a recurring weakness before changing your entire study plan. A rushed game, an unfamiliar opening and a long tournament game may expose different problems. Keep notes on the playing conditions so you can distinguish a repeated decision error from a bad evening. A label is useful only when you can point to the positions it describes.

To evaluate the loop, keep the old focus visible while you play the next serious games. Look for the decision you were supposed to improve, not only the final result. You can make a better defensive choice and still lose later; you can also win while repeating the same error. That distinction stops a weekly report becoming a substitute for reflection.

Families should check that the student understands the assignment without an adult translating every screen. If the daily practice is too difficult, too easy or disconnected from the lesson plan, adjust the routine. The reason to choose an integrated coaching loop is that it gets used between lessons. Device synchronisation alone does not establish that the practice suits the learner.

Chessvia’s Chessy is a voice-and-text option for discussing play and reviewing games. Its official page lists PGN input and Chess.com/Lichess import [15]. Start here: https://www.chessvia.ai/. Best for: a player who prefers talking through candidate moves to reading a report.

The site offers a free way to try Chessy and discusses paid access, so confirm the current plan amount and usage ceiling before paying [15]. Treat promotional cost comparisons as advertising, not as your checkout price. Check native mobile availability on your own device as well.

Test the difference between conversation and instruction. A lively response can keep you engaged, yet a useful coach must still address your actual question. Ask why your intended move fails and follow the answer on the board. Then ask about a different candidate move. The comparison should help you choose between plans, not simply restate that one move is stronger.

If you connect an online game history, choose which recent games represent the chess you want to improve. A collection of casual speed games may not tell the same story as the slower games you discuss in lessons. Keep the review goal explicit so a personalised response has a meaningful target.

For a child, an adult should preview the interaction and decide whether its tone, reading demands and account settings suit the family. Do not infer a minimum age or safeguarding guarantee from a colourful character. A child who cannot explain the question they asked may need a human-led board session before an open-ended conversational tool helps.

The practical buying test is whether you return to the same game and can make a better decision without prompting. If the interface is entertaining but you retain no usable idea, change the session design before paying for more interaction. A shorter, purposeful conversation can be more useful than letting the app narrate an entire game while you listen passively.

Use the following weighted buying rubric to compare the shortlist: training job, session format, free allowance, suitability for the learner and the point where human help is needed. These are suggested decision weights, not scores from an experiment. The ranked picks above reflect use cases, not measured learning gains.

The proposed weights put training fit first because a purchase should solve a named problem. Format and free allowance then determine whether the learner can practise in the available time and budget. Suitability and the need for human help complete the decision. Change the weights if access or supervision is your binding constraint; they are an editorial decision aid, not measured population data.

• What it trains (30): does the official site describe one primary job in precise verbs, or a vague mix of everything
• Session format (20): length and shape of a typical use, for example playing adaptive games in the browser [1], submitting a position for explanation [3], or drills generated from existing games [5]
• Free tier as stated (20): exactly what the vendor says is free, with currency and country as displayed where price appears
• Age fit (15): the reading level and supervision the vendor implies, not our estimate
• Where it stops (15): the explicit handoff point where a human coach must take over

Sourcing rule: product details come from each vendor's official pages as of 2 October 2026. A vendor's feature statement is not an independently tested efficacy claim. Where the table asks you to check an import route, mobile app or checkout amount, the vendor page does not settle it. That does not mean the feature is absent; it tells you what to confirm before choosing.

What we did not test: we did not run a controlled Stockfish agreement test, an illegal move count, or an Elo gain experiment across these apps, and we publish no engine agreement rates, hallucination counts, accuracy percentages, or Elo estimates in this guide. Where we describe hands on use, we label it as such; otherwise claims are sourced from official sites alone.

An AI coach can support practice and explanation, but a parent or coach should still judge whether the plan fits the learner’s circumstances. School workload, tournament nerves and motivation can change what is useful this week. Ask how the app’s advice connects to the next lesson rather than treating its output as the whole training plan.

The handoff is easy to spot in practice. An app has done its job when the student arrives at a lesson with material: games against Noctie [1], a weakness report from Aimchess [5], lines drilled on Chessable [8], a Lichess study full of positions they did not understand [10], or a Chess.com review they want to challenge [14]. The coach then turns that material into judgement: which weakness to fix first, which opening suits this student's style, what to say after a painful tournament. Apps generate evidence; coaches make decisions. That is why the advice throughout is one well matched app plus one human feedback loop, not a stack.

app.chesscore.io/games/aarav-vs-diya

GamesAarav R. vs. Diya K.

PASearch⌘K

AR

Aarav R. vs. Diya K.

Rapid · synced from Lichess · Tue 7:42 PM

Analyzed 2:04 AM

8

7

6

5

4

3

2

1a

b

c

d

e

f

g

h

14...Qe6? drops the knight

Eval swings +2.1 · the move to review first

Mistake

Missed fork on move 23

Same pattern as last Tuesday · drill it

Pattern

Endgame conversion was clean

Won vs. 1410 · French Defense

Best game

Add to homeworkOpen full analysis

Stockfish analysis finished · 2:04 AMEvery change logged · audit trail

Recommend one app per student, in writing, with a reason and a review date. Families trust a specific recommendation more than a list of six logos, and a built in expiry proves the academy is watching whether the tool works rather than collecting referral fees. A simple process any academy can run each term:

1. 1Name the student's single biggest bottleneck this term in one sentence, based on their recent games and lessons.
2. 2Map that bottleneck to one app from the two tables, and check the official site for current pricing on the day you recommend it, with currency and country as displayed.
3. 3Tell the parent exactly what usage looks like: sessions per week, minutes per session, and what the student should bring to the next lesson.
4. 4Review actual usage after four to six weeks, in a lesson, using the games or reports the app produced.
5. 5Keep or cancel: if the student is not opening the app, change the recommendation instead of blaming the student.

Record the recommendation somewhere the family can find it before renewal. State what the student should do, what evidence they should bring back and when the choice will be reconsidered. The communication channel matters less than the habit. A recommendation that disappears into a lobby conversation is difficult to evaluate later, especially when the parent and student remember different expectations.

app.chesscore.io/updates

ParentsSharma family

PASearch⌘K

Parent updates · this week

31 of 31 sent

SK

Sharma family

Receipt · March fee

MM

Mehta family

Weekly recap · Vihaan S.

KA

Khan family

Schedule change · Sat 4 PM

WhatsApp Business · connected

Weekly recap · Aarav R.

Aarav attended all 3 classes and won 2 arena games. His March report is in your portal.

Receipt · March fee

₹4,500 received via UPI AutoPay. Receipt saved.

9:42 AM

Sent after Coach Priya approved

Open the guardian portal

Magic link · nothing to install

Approved updates only · WhatsApp and parent portalEvery change logged · audit trail

Noctie is a sparring and guided-practice option, DecodeChess specialises in explanation, Aimchess in analytics, Chessable in course study and ChessKid in a child-focused playing environment. Lichess supplies a free analysis baseline. These are related training jobs, not interchangeable conversational coaches; use this section to avoid buying the wrong category.

Stockfish is the open source engine available in Lichess analysis [11][12]. The engine tells you which move is best; the training app wraps that output into something a student can act on the same afternoon. Five wrappers matter for a buying decision:

• Play: adaptive opponents that give realistic games, the job Noctie AI is built for [1]
• Explanation: turning engine output into plain language reasons and plans, the job DecodeChess is built for [3]
• Weakness analytics: mining a student's own online games for recurring mistakes, the job Aimchess is built for [5]
• Courses: opening and endgame material drilled by spaced repetition, the job Chessable is built for [8]
• Kids safety: a walled garden where young children play and learn without open chat, the job ChessKid is built for [6]

Lichess sits slightly apart. It is a full chess platform whose practice section, puzzle trainer, and studies cover several of those jobs at once, without charge [10][11]. It is less guided than the specialist apps, but it is the baseline every paid tool has to beat. This pillar is written from the buying seat: a parent choosing one tool for one child this term, or a coach recommending one per student.

Read the specialist tables by job. Noctie’s sparring and lessons [2], Aimchess’s reports [5] and DecodeChess’s explanations [4] overlap with parts of coaching but are not identical purchases. A cap-free browser engine does not supply the same interaction as a voice conversation. Choose the activity you need to complete, then check the allowance for that activity.

Match the app to the student's bottleneck this term, not to the vendor with the loudest claim. A seven year old beginner, a plateaued 1200 who plays daily online, and a teenager whose games collapse in the opening have three different problems, and the wrong app trains the wrong one. Name the one thing holding the student back, then pick the single tool built for that job.

For a young beginner, start with supervision and motivation rather than engine depth. ChessKid is the child-focused option in this guide [6]; preview its current lesson and account controls with the family. Lichess can be used for supervised puzzles without a subscription [11]. Neither choice should be based on an invented minimum age or a promise that the interface itself will keep the student engaged. Decide whether the child can use the activity with the support available at home.

For a club player who plays plenty but improves slowly, inspect the kind of feedback missing. Aimchess reports patterns across games [5]. Noctie combines practice with feedback and now lists lessons and scenarios [2]. DecodeChess explains positions [4]. A conversational coach from the shortlist may fit better when the unresolved question is a follow-up that a fixed report cannot answer. Ask the student to bring a recent position to the decision instead of assigning a tool from rating alone.

For a student whose games go wrong in the opening, decide whether the problem is forgetting a known line or misunderstanding the resulting position. Course drilling can address recall; game review can expose the decision that followed. Our Chessable alternatives guide separates scheduled repertoire trainers from shared study tools. For a family with no software budget, Lichess supplies a free workspace [11], and the coaching shortlist contains limited free interaction. Add a human review of the study plan where direction remains unclear.

Lichess is the free specialist baseline: it states that its features are free, while server analysis still has a daily usage cap [11]. Other products distinguish free trials, continuing free plans and paid access differently. Compare the actual activity you need, such as a conversation, a full-game review or a practice session, rather than treating every free label as the same offer.

As of 2 October 2026, the official pages publish Noctie’s EUR plans [2], DecodeChess’s free and paid decoding [4], and Aimchess’s free Standard and paid Premium offers [5]. The table records their units separately. A trial with full features is time-limited access, whereas a continuing free tier can carry a usage ceiling. Test the ceiling that will matter in your actual weekly routine.

The trap is not one subscription; it is stacking. A sensible budget is one paid app at a time, chosen against the bottleneck, reviewed after one term, and cancelled without guilt if the student stopped opening it. Our guide to free AI chess coaching options covers the zero budget version with Lichess and limited free tiers, and our Chessable alternatives page covers what to do when course style learning specifically is not landing.

For families in India, separate the global published price from the amount billed locally. The prices in this guide are global web prices, not local checkout quotes. Check currency, tax, renewal period and the specific language of instruction before recommending a subscription. A price in a comparison table is a starting point for that check, not a guarantee of the final payable amount.

Noctie's global pricing page displays EUR, while it says regional pricing may be available in the native app stores [2]. The web price is therefore not an India checkout quote. Check the final billing currency and any conversion charges with your payment provider rather than assuming the web display determines the app-store bill.

CircleChess is another provider to inspect when the family wants classroom learning as well as an app. Its official page links Caissa School of Chess, a ChessIQ report and a free WhatsApp starting point [16]. That does not establish a current course fee or prove it is the best fit for a particular city. Confirm which part of the offer is actually available to the student.

For India-based families, test the actual lesson and explanation language during the trial. A translated marketing page does not prove that the lesson audio, reports and support use the same language. Check the student’s device as well as the parent’s device before adopting the tool for homework.

Country availability, billing and teaching suitability are separate checks. A vendor's public page cannot tell you whether signup, imports or payment will work for your family. Follow the proposed study workflow on your own device before paying, and keep the final invoice currency and renewal period visible in your decision.