> ## Content Index
> Fetch the complete content index at: https://aiadoption.org/llms.txt
> Use this file to discover other available public pages before exploring further.

# The best SWE-bench Verified score reached 98.7% of the human baseline in 2025, up from 64.3%
- URL: https://aiadoption.org/data/ai-benchmarks-vs-human-baseline/
- Published: 2026-10-19T15:55:54.000Z
- Updated: 2026-10-21T06:02:35.000Z
- Description: On SWE-bench Verified, a software engineering test, the best AI score rose from 64.3% of the human baseline in 2024 to 98.7% in 2025. On OSWorld computer use it rose from 5.7% to 75.7%.
- Author: Manu
- Tags: Data, AI Benchmarks, Large Language Models, #chart, #ai-index-2026, #hub-benchmarks

98.7%

▲ 34.4 pts vs 2024

best SWE-bench Verified score in 2025, as a share of the human baseline

20242025

## What the chart shows

The AI Index scales each benchmark so that the best model of each year is shown as a share of the human baseline. On SWE-bench Verified, which tests autonomous software engineering, the best score rose from 64.3% of the human baseline in 2024 to 98.7% in 2025\. On OSWorld, which tests computer use, it rose from 5.7% to 75.7%.

Several reasoning benchmarks have passed the baseline: GPQA Diamond (PhD-level science) went from 47.8% in 2023 to 114.5% in 2025, and MMMU (multimodal reasoning) reached 106.1%. Older tests such as ImageNet (first passed in 2016), SQuAD 2.0 (2019) and SuperGLUE (2021) were surpassed earlier.

## Key numbers

| Measure                  | Value  |
| ------------------------ | ------ |
| SWE-bench Verified, 2024 | 64.3%  |
| SWE-bench Verified, 2025 | 98.7%  |
| OSWorld, 2024            | 5.7%   |
| OSWorld, 2025            | 75.7%  |
| GPQA Diamond, 2023       | 47.8%  |
| GPQA Diamond, 2025       | 114.5% |

## How to read it

- Each value is the best model’s score that year divided by the human baseline for that benchmark, as scaled by the AI Index; 100% means human-level performance.
- The chart shows the six benchmarks with a 2025 value. The data also covers five older ones (ImageNet, VQA, SQuAD 2.0, SuperGLUE, MMLU), all above the human baseline by their last year; download the CSV for all 11.
- The report text says SWE-bench Verified scores rose from about 60% in 2024; the data gives 64.3%, which we use.

**Source:** [AI Index](https://aiadoption.org/sources/#ai-index), 2026, via the AI Index Report 2026 (Stanford HAI), figure 2.1.1: *Select AI Index technical performance benchmarks vs. human performance*. Chart and text by AI Adoption.  
Numbers checked against the published data on 20 October 2026.  
[Download the data (CSV)](https://charts.aiadoption.org/ai-index/csv/ai-benchmarks-vs-human-baseline.csv?ref=aiadoption.org) · [Cite this chart](#source)