---
title: "Claude Opus 5 Benchmarks - RankedAGI"
description: "Detailed benchmark and metadata record for Claude Opus 5."
source: "https://rankedagi.com/models/claude-opus-5"
---

# Claude Opus 5

| Field | Value |
| --- | --- |
| Organization | Anthropic |
| License | Proprietary |
| Version | Latest |
| Released | 2026-07-24 |
| Input cost per million tokens | $5 |
| Output cost per million tokens | $25 |
| Last updated | 2026-07-24T17:45:19.784Z |

## Benchmarks

| Benchmark | Category | Value | Description | Source |
| --- | --- | --- | --- | --- |
| RankedAGI Coding | coding | 93.1% | RankedAGI Coding Score |  |
| RankedAGI Agentic | Agents | 93.8% | RankedAGI Agentic Score |  |
| BrowseComp | Agents | 90.8% | A benchmark for browsing agents | https://openai.com/index/browsecomp/ |
| RankedAGI Reasoning | reasoning | 87.1% | RankedAGI Reasoning Score |  |
| Humanity's Last Exam | reasoning | 56.3% | Multidisciplinary Reasoning (no tools) |  |
| Humanity's Last Exam w/ Tools | reasoning | 64.7% | Multidisciplinary Reasoning (with tools) |  |
| RankedAGI Math | math | 67.2% | RankedAGI Math Score |  |
| RankedAGI Overall | general | 77.7% | Overall RankedAGI score |  |
| GDPval-AA Elo | Agents | 1861 | Office Tasks (Artificial Analysis) |  |
| DeepSWE 1.1 | coding | 68.8% | Long-horizon engineering tasks, pass@1; mini-swe-agent harness run by Datacurve. | https://deepswe.datacurve.ai/ |

## Model Links

| Title | URL |
| --- | --- |
| Release | https://www.anthropic.com/news/claude-opus-5 |