---
title: "Claude Haiku 4.5 Benchmarks - RankedAGI"
description: "Detailed benchmark and metadata record for Claude Haiku 4.5."
source: "https://rankedagi.com/models/claude-haiku-4-5"
---

# Claude Haiku 4.5

| Field | Value |
| --- | --- |
| Organization | Anthropic |
| License | Proprietary |
| Version | Latest |
| Released | 2025-10-15 |
| Context window | 200K |
| Input cost per million tokens | $1 |
| Output cost per million tokens | $5 |
| Cached input cost per million tokens | $0.1 |
| Last updated | 2026-05-28T08:57:27.471Z |

## Benchmarks

| Benchmark | Category | Value | Description | Source |
| --- | --- | --- | --- | --- |
| RankedAGI Coding | coding | 54.6% | RankedAGI Coding Score |  |
| SWEBench Pro | coding | 39.5% | Diverse Agentic Coding |  |
| SWEBench Verified | coding | 73.3% | Agentic Coding | https://www.swebench.com/index.html |
| RankedAGI Agentic | Agents | 44.0% | RankedAGI Agentic Score |  |
| Code Arena | coding | 1316 | Code Arena Elo Score | https://arena.ai/leaderboard/code |
| Livebench Coding 26.1 | coding | 72.8% | Livebench Coding 26.01.08 | https://livebench.ai/#/?Coding=a&sort=Coding+Average&variants=true&highunseenbias=true |
| DeepSWE | coding | 0.0% | Measuring frontier coding agents on original, long-horizon engineering tasks | https://deepswe.datacurve.ai/ |
| LiveBench Agentic Coding 26.1 | coding | 41.7% |  | https://livebench.ai/#/?Agentic+Coding=a&sort=Agentic+Coding+Average&variants=true&highunseenbias=true |
| SvelteBench | coding | 84.4% | SvelteBench - Benchmark for Svelte | https://khromov.github.io/svelte-bench/benchmark-results-merged.html |
| RankedAGI Reasoning | reasoning | 46.5% | RankedAGI Reasoning Score |  |
| Humanity's Last Exam | reasoning | 9.7% | Multidisciplinary Reasoning (no tools) |  |
| GPQA Diamond | reasoning | 73.0% | Generalized Prefix Question Answering Score (Reasoning) PhD Level Reasoning |  |
| AIME 2025 | math | 96.3% | AIME 2025 Competition Math |  |
| Vending Bench 2 | Agents | $458.89 | 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 | imaging | 77.8% | Multimodal Understanding College-level visual problem-solving |  |
| RankedAGI Math | math | 68.4% | RankedAGI Math Score |  |
| RankedAGI Overall | general | 50.2% | Overall RankedAGI score |  |
| GDPval-AA Elo | Agents | 1177 | Office Tasks (Artificial Analysis) |  |
| Code DesignArena | Design | 1181 |  | https://www.designarena.ai/leaderboard/code |
| Toolathlon | Agents | 26.2% | Benchmark to assess language agents' general tool use in realistic environments. | https://toolathlon.xyz/introduction |

## Model Links

| Title | URL |
| --- | --- |
| Release | https://www.anthropic.com/news/claude-haiku-4-5 |