Benchfolio

Leaderboard / Model

GPT-5 Pro

OpenAI · Closed weights · Released Oct 7, 2025 · $15 in · $120 out per 1M tokens

n/aOverall · not enough data

Score by job

Coding

Not enough data · 0 of 5 benchmarks

Agents

Not enough data · 0 of 4 benchmarks

Reasoning

17 · rank 36 · 3 of 4 benchmarks

BenchmarkResultVariantz
ARC-AGI-218.3%gpt-5-pro-2025-10-06_unknown-2.10
SimpleBench61.6%gpt-5-pro-2025-10-06_high-0.38
Humanity’s Last Exam31.6%gpt-5-pro-2025-10-06_unknown-0.72

Math

Not enough data · 2 of 5 benchmarks

BenchmarkResultVariantz
FrontierMath Tiers 1–355.8%gpt-5-pro-2025-10-06_high-0.84
FrontierMath Tier 419.5%gpt-5-pro-2025-10-06_high-0.90

Writing

Not enough data · 0 of 2 benchmarks

Design

Not enough data · 0 of 1 benchmarks