AGIdex
agidex/Technologies/GPT-5.5
AI MODEL

GPT-5.5

CLOSED
RELEASED 2026-04

OpenAI's current flagship. Leads ARC-AGI-2 at 85%. Strong on coding and reasoning.

Benchmark coverage · per sub-capability

Each score is anchored to human baseline (100). Source URLs link to the original benchmark, leaderboard, or release note.

№ 04ReasoningCOGNITIVE
Sub-capabilityQualitySourceScore vs. baselineScore
Abstract Reasoningindependentarcprize.org/leaderboard
85%
Logical Reasoningvendor reportedopenai.com/index/introducing-gpt-5-5
76%
№ 01CommunicationCOGNITIVE
Sub-capabilityQualitySourceScore vs. baselineScore
Language Generationindependentlmarena.ai
104%
№ 07CreativityCOGNITIVE
Sub-capabilityQualitySourceScore vs. baselineScore
Open-ended Problem Solvingvendor reportedopenai.com/index/introducing-gpt-5-5
52%