GPT-5.6 Terra Benchmark Details

GPT-5.6 Terra currently shows benchmark results led by ARC-AGI (2 / 68, score 96.50), ARC-AGI-2 (7 / 62, score 83.90), DeepSWE (3 / 19, score 69.60).

Benchmark Results

GPT-5.6 Terra

Benchmark Results

General Knowledge

5 evaluations

Benchmark / mode

Score

Rank/total

ARC-AGI

Thinking Level · Max

96.50

2 / 68

ARC-AGI-2

Thinking Level · Max

83.90

7 / 62

Vals Index

Thinking Level · Extra High

65.14

2 / 2

AA Intelligence Index

Thinking Level · Max

2 / 3

ARC-AGI-3

Thinking Level · Max

0.80

2 / 9

AI Agent - Tool Usage

2 evaluations

Benchmark / mode

Score

Rank/total

TerminalBench 2.1

Thinking Level · Max

87.40

5 / 27

Agents' Last Exam

Thinking Level · Extra HighTools

50.40

2 / 3

Coding and Software Engineer

2 evaluations

Benchmark / mode

Score

Rank/total

AA Coding Agent Index

Thinking Level · Extra HighTools

77.40

2 / 3

DeepSWE

Thinking Level · Extra HighTools

69.60

3 / 19

Compare with other models