OpenAI

GPT-5

Announced Aug 7, 2025

Fast answers and deeper reasoning

What could it do?

It switched between quick answers and deeper reasoning, with better coding, writing, and health answers, fewer made up facts, and less flattery.

What changed?

A unified system with routing between fast answers and deeper reasoning.

WHY IT MATTERED

Adaptive reasoning

Introduced a new default generation in ChatGPT.

Which version is this?

A routed product and an individual API model are not interchangeable benchmark entries.

Research draft

Capabilities describe the developer’s announcement. This entry has not completed source review.

Read original source

Benchmark results

EPOCH AI CAPABILITY ESTIMATE

150.0index points

Tested variant: GPT-5Source interval not suppliedVariant date in source: Aug 7, 2025

A benchmark estimate, not a percentage or capability multiplier. Reasoning settings are not specified in this source table. Historical estimates can change in later snapshots.

Epoch AI methodology ↗Download the source snapshotChecked Oct 7, 2026 · CC BY 4.0

PUBLISHED BENCHMARK RESULT

1435rating points

Tested: gpt-5-highText Arena Overall · October 8, 2026Reported interval: 1430 to 144031,723 votes

One explicitly named variant per release. Scores come from the same Overall snapshot; preliminary entries and reported intervals are preserved. These are current ratings of earlier variants, not their launch day ratings.

Source: Text Arena ↗Download selected resultsChecked Oct 8, 2026

Understand the benchmarks Find this release on the chart