---
title: "Claude Opus 4.6 Benchmarks - RankedAGI"
description: "Detailed benchmark and metadata record for Claude Opus 4.6."
source: "https://rankedagi.com/models/claude-opus-4-6"
---

# Claude Opus 4.6

| Field | Value |
| --- | --- |
| Organization | Anthropic |
| License | Proprietary |
| Type | Thinking |
| Version | Latest |
| Released | 2026-02-05 |
| Context window | 1M |
| Input cost per million tokens | $5 |
| Output cost per million tokens | $25 |
| Cached input cost per million tokens | $0.5 |
| Model ID | claude-opus-4-6 |
| Last updated | 2026-04-27T16:52:03.689Z |

## Benchmarks

| Benchmark | Category | Value | Description | Source |
| --- | --- | --- | --- | --- |
| RankedAGI Coding | coding | 79.8% | RankedAGI Coding Score |  |
| SWEBench Pro | coding | 57.3% | Diverse Agentic Coding |  |
| SWEBench Multilingual | coding | 77.8% |  |  |
| SWEBench Verified | coding | 80.8% | Agentic Coding | https://www.swebench.com/index.html |
| Terminal-Bench 2.0 | coding | 65.4% | Agentic Terminal Coding | https://www.tbench.ai/leaderboard/terminal-bench/2.0 |
| RankedAGI Agentic | Agents | 68.6% | RankedAGI Agentic Score |  |
| BrowseComp | Agents | 84.0% | A benchmark for browsing agents | https://openai.com/index/browsecomp/ |
| Code Arena | coding | 1551 | Code Arena Elo Score | https://arena.ai/leaderboard/code |
| Livebench Coding 26.1 | coding | 78.2% | Livebench Coding 26.01.08 | https://livebench.ai/#/?Coding=a&sort=Coding+Average&variants=true&highunseenbias=true |
| LiveBench Agentic Coding 26.1 | coding | 61.7% |  | https://livebench.ai/#/?Agentic+Coding=a&sort=Agentic+Coding+Average&variants=true&highunseenbias=true |
| OSWorld-Verified | Agents | 72.7% |  |  |
| SvelteBench | coding | 100.0% | SvelteBench - Benchmark for Svelte | https://khromov.github.io/svelte-bench/benchmark-results-merged.html |
| LiveCodeBench v6 | coding | 88.8% | LiveCodeBench v6 - Evaluation of Large Language Models for Code |  |
| CyberGym | safety | 66.6% |  |  |
| Codeforces ELO Score | coding | 2352 | Codeforces ELO Score |  |
| RankedAGI Reasoning | reasoning | 76.8% | RankedAGI Reasoning Score |  |
| Humanity's Last Exam | reasoning | 40.0% | Multidisciplinary Reasoning (no tools) |  |
| Humanity's Last Exam w/ Tools | reasoning | 53.0% | Multidisciplinary Reasoning (with tools) |  |
| GPQA Diamond | reasoning | 91.3% | Generalized Prefix Question Answering Score (Reasoning) PhD Level Reasoning |  |
| Text Arena | general | 1503 | ChatArena (LMSYS) ELO Score | https://arena.ai/leaderboard/text |
| AIME 2026 | math | 96.7% | AIME 2026 Competition Math | https://matharena.ai/?comp=aime--aime_2026&view=problem |
| Vending Bench 2 | Agents | $8,017.59 | 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 |
| NYT Connections Extended | reasoning | 94.7% | NYT Connections Extended Version | https://github.com/lechmazur/nyt-connections/ |
| MMLU Pro | knowledge | 89.1% | Graduate-level visual problem-solving |  |
| MMMU-Pro | imaging | 73.9% | Multimodal Understanding |  |
| MMMU Pro w/Tools | imaging | 77.3% |  |  |
| RankedAGI Math | math | 76.8% | RankedAGI Math Score |  |
| RankedAGI Overall | general | 68.1% | Overall RankedAGI score |  |
| ARC AGI 2 | reasoning | 68.8% | Abstract Reasoning Puzzles (Public) |  |
| LiveCodeBench Pro | coding | 71.7% | Competitive Coding |  |
| 𝜏²-Bench Telecom | Agents | 99.3% | Agentic Tool Use |  |
| 𝜏²-Bench Retail | Agents | 91.9% | Agentic Tool Use |  |
| HealthBench Hard | Health | 14.8% | Open-ended Health Queries |  |
| MedXpertQA Text | Health | 52.1% | Medical Multiple Choice |  |
| MedXpertQA (MM) | Health | 64.8% | Medical Multiple Choice |  |
| DeepSearch QA | Agents | 73.7% | Agentic Search |  |
| GDPval-AA Elo | Agents | 1619 | Office Tasks (Artificial Analysis) |  |
| MCP Atlas | Agents | 59.5% | Scaled Tool Use |  |
| Finance Agent | Agents | 60.7% | Agentic Financial Analysis |  |
| MMMLU | Uncategorized | 91.1% | Multilingual Q&A |  |
| CharXiv Reasoning | Multimodal | 69.1% | Figure Understanding |  |
| CharXiv Reasoning w/ Tools | Multimodal | 84.7% | Figure Understanding |  |
| Code DesignArena | Design | 1354 |  | https://www.designarena.ai/leaderboard/code |
| Toolathlon | Agents | 47.2% | Benchmark to assess language agents' general tool use in realistic environments. | https://toolathlon.xyz/introduction |
| Tau3 Bench | Agents | 72.4% |  | https://sierra.ai/resources/research/tau-3-bench |

## Model Links

| Title | URL |
| --- | --- |
| Release | https://www.anthropic.com/news/claude-opus-4-6 |