SWE-Lancer (Diamond): leaderboard
Tests AI agents on 1,400+ real freelance software engineering tasks sourced from Upwork, ranging from $50 bug fixes to $32,000 feature implementations, graded with end-to-end tests.
Metric: Score (%). Source: swelancer.github.io. Saturation forecast: Around January 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.3 Codex | 81.4 | #69 |
| 2 | GPT-5.2 | 74.6 | #110 |
| 3 | GPT-4.5 | 17.4 | #274 |
| 4 | GPT-4o | 12.4 | #343 |
| 5 | O3 Mini | 7.4 | #272 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=swe-lancer-diamond · How It Works · Data refreshed daily, snapshot 2026-10-20.