Model

GPT-5.5 (High)

CLOSED

OpenAI · GPT-5.5 family · released Apr 23, 2026

Reasoning-effort sibling of the already-catalogued base 'GPT-5.5' entry. Same weights, same pricing/context as the default/xhigh siblings; distinguished only by the reasoning.effort=high API parameter (also selectable in the ChatGPT model picker's effort selector). Same-family, same-release-date, same-paramsB per the effort-tier convention.

ReasoningCodingVisionFunction callingTool useAgentic

2560.2

Elo · rank #27

Parameters

Undisclosed

Active params

—

Context

1.05M tokens

Architecture

Undisclosed architecture (presumed dense transformer, reasoning-tuned); codename "Spud"; 'high' reasoning-effort configuration

License

Proprietary (OpenAI API Terms)

Languages

—

API price (in/out)

$5 / $30

Modalities

text · vision

Benchmark results

Bar shows position within the tracked field; marker = field best

Agents' Last ExamAgents44.0%#9

best: GPT-6 Astra · 59.3%

GDPval-AAAgents1472#25

best: Claude Fable 5 · 1932

GPQA DiamondReasoning93.2%#11

best: GPT-6 Astra · 96.0%

Humanity's Last ExamReasoning43.0%#21

best: Claude Opus 5 · 64.7%

IFBenchReasoning71.6%#30

best: MiniMax M3 · 83.0%

MMMU-ProVision81.1%#8

best: Claude Opus 4.7 · 85.5%

Terminal-Bench 2.0Coding79.4%#18

best: Gemini 3.8 Flash · 89.4%

Run it locally

Closed weights — available via API only. No local deployment.

Input / M tok

Output / M tok

$30

GPT-5.5 family

Elo progression across releases

GPT-5.5Apr 20262634.5 GPT-5.5 (High)Apr 20262560.2 GPT-5.5 (xHigh)Apr 20262677.5

Resources

DOCShttps://developers.openai.com/api/docs/models/gpt-5.5 APIAvailable via OpenAI API

API price $5/$30 · each benchmark row carries its own source badge (see methodology)