OpenAIclosed weightsReasoning
GPT-5.5
First OpenAI model past 80% on Terminal-Bench 2.0; 2x price above 272K input.
Apr 23, 2026
$5 / $30
$11
1M
Benchmark scores
| Benchmark | Score | Δ vs prior | Rank (current) | Configuration | Source |
|---|---|---|---|---|---|
Terminal-Bench 2.1 | 76.4% | #8 of 21 | leaderboard | benchlm.ai ↗ | |
GPQA Diamond | 93.6% | #3 of 11 | leaderboard | benchlm.ai ↗ | |
Humanity's Last Exam | 41.4% | #10 of 15 | no tools · lab-reported | vellum.ai ↗ | |
ARC-AGI-2 | 85% | #6 of 10 | leaderboard | benchlm.ai ↗ | |
τ²-bench (Telecom) | 98% | #1 of 10 | leaderboard | benchlm.ai ↗ | |
Artificial Analysis Intelligence Index | 54.8 | #2 of 11 | leaderboard | benchlm.ai ↗ | |
LMArena Text | 1481 | #1 of 5 | high, Jul 2026 · leaderboard | toolcenter.ai ↗ |