GPT-5.5

OpenAI·Operational·Released 2026-04-23textvision

Context1,000k

Max output128k

Input / 1M$5 ↗

Output / 1M$30 ↗

Cached in$0.5

Benchmarks21

Benchmarks

Cited public sources · rank is exact-benchmark, apples-to-apples

AIME 2026 (MathArena)

#1/21100% ↗

FrontierMath Tier 4 v2 (Epoch AI run)

#6/4472.5% ↗

FrontierMath Tiers 1-3 v2 (Epoch AI run)

#7/5985.3% ↗

GPQA Diamond (Epoch AI run)

#24/7790.7% ↗

OTIS Mock AIME 2024-2025 (Epoch AI run)

#47/7684.4% ↗

SimpleQA Verified (Epoch AI run)

#10/5363% ↗

LMArena Elo

#7/621465.6 ↗

ARC-AGI-1 (Verified)

#1/495% ↗

ARC-AGI-2 (Verified)

#1/785% ↗

BrowseComp

#3/1584.4% ↗

FrontierMath Tier 1-3

#2/351.7% ↗

FrontierMath Tier 4

#3/335.4% ↗

GDPval (wins or ties)

#1/684.9% ↗

MCP Atlas

#3/975.3% ↗

MMMU-Pro (no tools)

#2/681.2% ↗

MMMU-Pro (with tools)

#1/783.2% ↗

OSWorld-Verified

#5/1278.7% ↗

SWE-Bench Pro

#9/1558.6% ↗

Tau2-bench Telecom

#3/798% ↗

Terminal-Bench 2.0

#1/1382.7% ↗

Toolathlon

#1/655.6% ↗

Compare

vs GPT-6 Astra vs GPT-5.6 Sol vs GPT-5.6 Terra vs GPT-5.6 Luna vs Claude Fable 5.1 vs Claude Opus 5 vs Claude Sonnet 5 vs Claude Haiku 4.5

More from OpenAI

GPT-6 Astra GPT-5.6 Sol GPT-5.6 Terra GPT-5.6 Luna GPT-5.5 Pro GPT-5.4

Access GPT-5.5 and every other model through one endpoint with automatic failover: Respan gateway.