Skip to content
AI Atlas
Familyroot · GPT

GPT 5

All 11 models in the terminal →

first seen 11 Sept 2026

family_01M2A55TRRTCETMKZ529F3R0CQ

Models
11
First release
7 Aug 2025
Latest release
16 Oct 2025
Parameters
—
Artifacts
0
Providers
1

Releases

Family releases by statusActiveDeprecatedSept 25Oct 25gpt-5 — 7 Aug 2025gpt-5-mini — 7 Aug 2025gpt-5-nano — 7 Aug 2025gpt-5-pro — 6 Oct 2025GPT-5 Image — 14 Oct 2025GPT-5 Image Mini — 16 Oct 2025

6 dated releases · dot size = best benchmark rank of the member · 5 members without a sourced release date not shown

Members 11

Family members
ModelParamsContextOpennessLicenceReleasedStatusBest rankCompare
GPT-5 Image MiniOpenAI—400K——16 Oct 2025Active—
GPT-5 ImageOpenAI—400K——14 Oct 2025Active—
gpt-5-proOpenAI—400KClosed—6 Oct 2025Deprecated—
gpt-5OpenAI—400KClosed—7 Aug 2025Deprecated#1 Aider polyglot
gpt-5-miniOpenAI—400KClosed—7 Aug 2025Deprecated#13 SWE-bench Multilingual
gpt-5-nanoOpenAI—400KClosed—7 Aug 2025Deprecated#34 SWE-bench Verified
GPT-5 ChatOpenAI——Closed——Retired—
gpt-5-chatgptOpenAI—128KClosed——Deprecated#152 Terminal-Bench
GPT-5-CodexOpenAI—400KClosed——Retired#33 IFBench
gpt-5-codex-miniOpenAI—————Active—
gpt-5-search-apiOpenAI—————Active—

Lineage among members 0 relations

Open in Graph →

No lineage relation recorded among these members.

Benchmark progress

Benchmark progress needs at least two dated members with a rank on the same benchmark. Ranks (not scores) are what the family endpoint provides.

Aggregates are computed from the members' sourced claims; members without a release date or parameter count are left out of the range and timeline rather than guessed. Methodology →