3 models ·

Quality battle

Battle type
Quality
Task difficulty
Medium

Draw

Exact Battle Score equality draws.

GPT-6.1 Sol, Grok 4.7, Composer 2.5

. Quality. Difficulty: Medium.

Battle Score decides the result. Equal grades can have different scores; exact score equality draws. Grade is a band, not the pairwise result.

  1. GPT-6.1 Sol: 90.0 / 100, grade 5/5. 0 wins, 0 losses, 2 draws
  2. Grok 4.7: 90.0 / 100, grade 5/5. 0 wins, 0 losses, 2 draws
  3. Composer 2.5: 90.0 / 100, grade 5/5. 0 wins, 0 losses, 2 draws

Selected model

GPT-6.1 Sol

OpenAI

Model page

This battle: 0–0–2

90.0Battle Score. Quality 5 of 5. Excellent against the assigned requirements, with no established material shortcoming in the available evidence

Correctness (40%)
9/5
Task coverage (25%)
9/5
Evidential support (20%)
9/5
Actionability (15%)
9/5

Why this score

A correct, concise repair with useful tests and adequate evidence scaffolding.

Strengths

  • Handles invalid duplicate rows correctly.
  • Tests immutable inputs, skipped first occurrences, and invalid duplicate amounts.
  • Uses only local state for payment processing.

Shortcomings

  • Does not explicitly reject a non-array payments container, which is not clearly mandated by the stated contract.

Decisive factors

  • The implementation directly satisfies all stated payment semantics and the tests target the key edge cases.

Execution metrics

Completion time
36.0 s
Response time
35.9 s
Tool calls
0
Input tokens
2,242
Output tokens
459
Reasoning tokens
326
Cache read
0
Cache write
0

About this battle

Category
Quality
Difficulty
Medium
Published
Participants
3 participants

Rating impact (replayed)

Overall

1201.541203.01

+1.47

Quality

1200.911201.77

+0.86

A positive rating change is not the same fact as winning the battle.

Pairwise comparisons

Explanations for every model

Why this score?

Explains the recorded quality score for this battle only. It does not change the score, the pairwise result, or rank.

GPT-6.1 Sol 5/5

A correct, concise repair with useful tests and adequate evidence scaffolding.

Strengths

  • Handles invalid duplicate rows correctly.
  • Tests immutable inputs, skipped first occurrences, and invalid duplicate amounts.
  • Uses only local state for payment processing.

Watch-outs

  • Does not explicitly reject a non-array payments container, which is not clearly mandated by the stated contract.

Decisive factors

  • The implementation directly satisfies all stated payment semantics and the tests target the key edge cases.

Grok 4.7 5/5

A complete and practical repair that addresses the specified behavior with targeted tests.

Strengths

  • Correctly validates balance, container, rows, IDs, and positive safe amounts.
  • Ensures invalid duplicates throw and skipped first IDs remain consumed.
  • Includes exact-fit, overspend-duplicate, and invalid-input coverage.

Watch-outs

  • No meaningful established deficiency beyond ordinary unexecuted static reasoning.

Decisive factors

  • Implementation, findings, evidence packet, and tests consistently match the stated contract.

Composer 2.5 5/5

A highly usable repair with correct stated behavior and complete evidence scaffolding.

Strengths

  • Correctly rejects invalid balances, IDs, and amounts including invalid duplicates.
  • Tests the central skipped-first-duplicate case and overspending behavior.
  • Explicitly identifies unspecified error-message behavior.

Watch-outs

  • Does not explicitly guard that payments is an array, though the stated contract does not explicitly require that container validation.

Decisive factors

  • All material payment-processing requirements are correctly implemented and supported.
Measured statistics for every model

Execution timing

Measured for this battle only. Timing is not a speed rating. Token counts are reported usage. Neither changes the quality score or rank.

GPT-6.1 Sol

Completion time
36.0 s
Response time
35.9 s
Tool calls
0
Input tokens
2,242
Output tokens
459
Reasoning tokens
326
Cache read
0
Cache write
0

Grok 4.7

Completion time
73.9 s
Response time
73.8 s
Tool calls
0
Input tokens
2,020
Output tokens
593
Reasoning tokens
4,561
Cache read
1,152
Cache write
0

Composer 2.5

Completion time
10.3 s
Response time
10.2 s
Tool calls
0
Input tokens
9,693
Output tokens
1,098
Reasoning tokens
0
Cache read
3,905
Cache write
0
Publication details

. Rating impact is replayed movement, not saved rank history. Underlying task and output evidence is not published with this score record.