1. Home
  2. /
  3. Claude Haiku 4.5

Benchmark page

Claude Haiku 4.5 chess benchmark results

Claude Haiku 4.5 scored 8.7% strict accuracy on ChessBench, solving 13 of 150 fixed Lichess tactics when the answer had to match the expected UCI move line exactly.

Strict accuracy

8.7%

Correct puzzles

13 / 150

Prompt tokens

76,316

Total cost

$0.15

Benchmark updated

Mar 24, 2026, 5:38 AM

Tactical breakdown

Fork

13.3%

Find the tactical line that creates a fork.

Hanging piece

0.0%

Find the line that wins an undefended piece.

Mate in 1

30.0%

Solve a forced mate in one move.

Mate in 2

0.0%

Solve a forced mate in two moves.

Pin

0.0%

Find the tactical line that exploits a pin.

Why this page can rank

This page gives Google a stable, canonical URL for the model and describes the benchmark result in plain HTML instead of only inside the interactive explorer.

It also links back to the benchmark methodology and exposes model-specific copy, numbers, and sample puzzle outcomes that are relevant to searches like "Claude Haiku 4.5chess benchmark" and "Claude Haiku 4.5tactical reasoning results".

For the full dataset definition and scoring details, visit the ChessBench dataset and methodology page.

Sample benchmark outcomes

These examples add readable, model-specific context beyond a single leaderboard number.

Open the full explorer

Mate in 1

Puzzle 5ZUAJ

Solved

Expected line: g2c2

Model line: g2c2

Parse status: ok

View original Lichess puzzle

Mate in 1

Puzzle lCsaK

Solved

Expected line: d5f7

Model line: d5f7

Parse status: ok

View original Lichess puzzle

Mate in 1

Puzzle T6Isu

Solved

Expected line: d4e2

Model line: d4e2

Parse status: san_converted

View original Lichess puzzle

Mate in 1

Puzzle Gxali

Missed

Expected line: b5e8

Model line: c5e6

Parse status: san_converted

View original Lichess puzzle

Mate in 1

Puzzle k82EC

Missed

Expected line: b1h7

Model line: No parseable line returned

Parse status: invalid_count

View original Lichess puzzle

Mate in 1

Puzzle LCT4l

Missed

Expected line: d8f8

Model line: f1f7

Parse status: loose_invalid_count

View original Lichess puzzle