Analyze FEN positions, query verified single-solution tactical puzzles by motif, and run batch PGN game reviews over HTTP.
Three HTTP endpoints for position analysis, tactical puzzle queries, and PGN game review.
Pass a FEN string and optional candidate move to retrieve centipawn scores, win/draw/loss probabilities, principal variations, and move commentary.
Filter single-solution tactical positions by motif tag and rating range. Every position is checked to ensure only one winning move exists.
Submit PGN game strings to identify blunders, missed tactics, and accuracy breakdowns across opening, middlegame, and endgame phases.
| Endpoint | Description | Typical Response | Fields Returned |
|---|---|---|---|
| POST /v1/evaluate | Position evaluation and Multi-PV lines | ~18 ms | eval_cp, wdl, best_move, pv, coaching_summary |
| GET /v1/puzzles | Tactical and mating pattern query | ~7 ms | puzzle_id, fen, best_move, motif, difficulty_elo |
| POST /v1/reports | PGN game accuracy and blunder log | ~340 ms / game | accuracy, critical_moments, missed_tactics |
Use the Python client or standard HTTP requests. Test tokens are rate-limited to 100 requests per hour.
Use this token for local testing or generate a new one below.
plyvero-api.somer-5b7.workers.devfrom plyvero import PlyveroClient
client = PlyveroClient(api_key="ply_sbx_live_94a8f10c3e7b2d4190a2")
# Evaluate a position
analysis = client.evaluate(
fen="r1bqk2r/pppp1ppp/2n2n2/2b1p3/2B1P3/5N2/PPPP1PPP/RNBQK2R w KQkq - 4 4",
played_move="d2d3",
multipv=3,
include_coaching=True
)
print(analysis.eval_cp, analysis.coaching_summary)
# Fetch tactical puzzles by motif
puzzles = client.puzzles.query(
motif="back_rank_deflection",
rating_min=1900,
rating_max=2250,
limit=5
)
Monthly plans based on evaluation volume.
For personal projects and testing.
For coaches and training clubs.
For chess websites and apps.
Plyvero Labs is an independent software project built by Saksham Omer, focused on chess position evaluation, curated tactical datasets, and automated game review tools.
Evaluation requests are processed by stateless Linux worker containers backed by object storage for PGN archives and low-latency REST endpoints.
Maintainer: Saksham Omer
Email: Somer@student.gitam.edu
GitHub: github.com/somer-eng