---
title: "GPT 5.5 Benchmarks - RankedAGI"
description: "Detailed benchmark and metadata record for GPT 5.5."
source: "https://rankedagi.com/models/gpt-5-5"
---

# GPT 5.5

| Field | Value |
| --- | --- |
| Organization | OpenAI |
| License | Proprietary |
| Type | Thinking |
| Version | Latest |
| Released | 2026-04-23 |
| Context window | 1M |
| Knowledge cutoff | 2025-12-01 |
| Input cost per million tokens | $5 |
| Output cost per million tokens | $30 |
| Cached input cost per million tokens | $0.5 |
| Model ID | gpt-5.5 |
| Last updated | 2026-05-28T21:32:30.135Z |

## Benchmarks

| Benchmark | Category | Value | Description | Source |
| --- | --- | --- | --- | --- |
| RankedAGI Coding | coding | 86.7% | RankedAGI Coding Score |  |
| SWEBench Pro | coding | 58.6% | Diverse Agentic Coding |  |
| Terminal-Bench 2.0 | coding | 82.7% | Agentic Terminal Coding | https://www.tbench.ai/leaderboard/terminal-bench/2.0 |
| Terminal-Bench 2.1 | coding | 83.4% | Agentic Terminal Coding | https://www.tbench.ai/leaderboard/terminal-bench/2.1 |
| RankedAGI Agentic | Agents | 77.5% | RankedAGI Agentic Score |  |
| BrowseComp | Agents | 84.4% | A benchmark for browsing agents | https://openai.com/index/browsecomp/ |
| Code Arena | coding | 1500 | Code Arena Elo Score | https://arena.ai/leaderboard/code |
| Livebench Coding 26.1 | coding | 82.5% | Livebench Coding 26.01.08 | https://livebench.ai/#/?Coding=a&sort=Coding+Average&variants=true&highunseenbias=true |
| DeepSWE | coding | 70.0% | Measuring frontier coding agents on original, long-horizon engineering tasks | https://deepswe.datacurve.ai/ |
| LiveBench Agentic Coding 26.1 | coding | 56.7% |  | https://livebench.ai/#/?Agentic+Coding=a&sort=Agentic+Coding+Average&variants=true&highunseenbias=true |
| OSWorld-Verified | Agents | 78.7% |  |  |
| SvelteBench | coding | 97.8% | SvelteBench - Benchmark for Svelte | https://khromov.github.io/svelte-bench/benchmark-results-merged.html |
| CyberGym | safety | 81.8% |  |  |
| RankedAGI Reasoning | reasoning | 85.1% | RankedAGI Reasoning Score |  |
| Humanity's Last Exam | reasoning | 41.4% | Multidisciplinary Reasoning (no tools) |  |
| Humanity's Last Exam w/ Tools | reasoning | 52.2% | Multidisciplinary Reasoning (with tools) |  |
| GPQA Diamond | reasoning | 93.6% | Generalized Prefix Question Answering Score (Reasoning) PhD Level Reasoning |  |
| Text Arena | general | 1488 | ChatArena (LMSYS) ELO Score | https://arena.ai/leaderboard/text |
| Vending Bench 2 | Agents | $7,523.84 | Benchmark for measuring AI model performance on running a business over long time horizons. Models are tasked with running a simulated vending machine business over a year and scored on their bank account balance at the end. | https://andonlabs.com/evals/vending-bench-2 |
| MMMU-Pro | imaging | 81.2% | Multimodal Understanding |  |
| MMMU Pro w/Tools | imaging | 83.2% |  |  |
| RankedAGI Math | math | 72.2% | RankedAGI Math Score |  |
| RankedAGI Overall | general | 72.8% | Overall RankedAGI score |  |
| ARC AGI 2 | reasoning | 85.0% | Abstract Reasoning Puzzles (Public) |  |
| 𝜏²-Bench Telecom | Agents | 98.0% | Agentic Tool Use |  |
| MCP Atlas | Agents | 75.3% | Scaled Tool Use |  |
| Finance Agent | Agents | 60.0% | Agentic Financial Analysis |  |
| Code DesignArena | Design | 1313 |  | https://www.designarena.ai/leaderboard/code |
| Toolathlon | Agents | 55.6% | Benchmark to assess language agents' general tool use in realistic environments. | https://toolathlon.xyz/introduction |
| Cursor Bench 3.1 | coding | 64.3% | Benchmark to evaluate agents on ambiguous, multi-file tasks from real Cursor sessions. | https://cursor.com/evals |

## Model Links

| Title | URL |
| --- | --- |
| Release | https://openai.com/index/introducing-gpt-5-5/ |