SWE-Lancer: leaderboard
Real freelance software engineering tasks from Upwork valued at $1M total - from $50 bug fixes to $32K feature implementations. Graded via end-to-end tests. Diamond subset is triple-verified by senior engineers.
Metric: Score (%). Source: openai.com. Source: www.vals.ai. Saturation forecast: Around January 2027. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.1 Codex | 66.3 | #159 |
| 2 | GPT-4.5 | 37.3 | #274 |
| 3 | GPT-4o | 32.6 | #343 |
| 4 | O3 Mini | 18 | #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 · How It Works · Data refreshed daily, snapshot 2026-10-20.