GPT-5.5 (low)

GPT-5.5 (low)

OpenAI · Released Apr 23, 2026 Professional
Intelligence #14 / 590
82.0 Our Score
Speed #182 / 278
66.4 tokens / sec
Input #549 / 590
$5.00 per 1M tokens
Output #558 / 590
$30.00 per 1M tokens
Context
— Not reported

Analysis Summary

GPT-5.5 (low) is the minimum reasoning-budget variant of OpenAI's GPT-5.5 family, with an intelligence index of 43.5, coding index of 60.9, and agentic index of 68.1. Its instruction-following score of 0.644, long-context retrieval of 0.720, and terminal benchmark of 0.523 are strong across the board, and its tau2 of 0.839 indicates reliable structured task completion.

For businesses, this variant offers a compelling balance: near-frontier capability at the same $5/$30 per million token price as the medium and non-reasoning variants, with lower latency due to reduced reasoning overhead. It suits coding assistance, agentic workflows, and content tasks where speed matters more than maximum reasoning depth.

Teams running high-throughput pipelines or latency-sensitive applications should consider this variant as a primary option, escalating to the medium or high variants only when task complexity demands it.

Assessed June 30, 2026

Editorial notes

GPT-5.5 (low) from OpenAI delivers a 43.5 intelligence index, 60.9 coding index, and 68.1 agentic index with strong instruction following and long-context retrieval, offering near-high-tier performance at the standard $5/$30 price point.

Rankings consider pricing, capabilities, benchmarks, and real-world applicability and are refreshed as new models launch. Feedback?

Performance Profile

Intelligence6.6Technical8.8Value4.5Content7.8
Intelligence 6.6/10
Technical 8.8/10
Content 7.8/10
Value 4.5/10

How GPT-5.5 (low) compares

GPT-5.5 (low) ranks #19 of 385 AI models we track for overall intelligence, #14 of 129 for coding, #27 of 293 for agentic tasks. At $5.00 per million input tokens it is cheaper than 7% of comparable models.

Performance Indices

Source: Artificial Analysis

43.5 Intelligence Index
60.9 Coding Index
68.1 Agentic Index

Benchmark Scores

Intelligence

GPQA Diamond 91% Graduate-level scientific reasoning
HLE 31% Humanity's Last Exam
SciCode 51.6% Scientific computing

Technical

TerminalBench Hard 52.3% Agentic terminal tasks
τ²-Bench 83.9% Conversational agent benchmark

Content

IFBench 64.4% Instruction following
LCR 72% Long-context reasoning

Benchmark data from Artificial Analysis and Hugging Face

How does GPT-5.5 (low) stack up?

Compare side-by-side with other professional models.

Compare Models

Model Information

ProviderOpenAI
Release Date April 23, 2026
Status Active

Pricing

Token Type Cost per 1M tokens Cost per 1K tokens
Input $5.00 $0.005000
Output $30.00 $0.030000

Leaderboard Categories

Frequently asked questions about GPT-5.5 (low)

How much does GPT-5.5 (low) cost?

GPT-5.5 (low) costs $5.00 per million input tokens and $30.00 per million output tokens.

Is GPT-5.5 (low) good for coding?

On our coding benchmark index, GPT-5.5 (low) ranks #14 of 129 models, placing it in the top quartile of the field for code generation and debugging.

Who created GPT-5.5 (low)?

GPT-5.5 (low) is developed by OpenAI and was released on April 23, 2026.