reasoning benchmark
SWE-Lancer (IC-Diamond subset) is a benchmark of real-world freelance software engineering tasks from Upwork, ranging from $50 bug fixes to $32,000 feature implementations. It evaluates AI models on independent engineering tasks using end-to-end tests triple-verified by experienced software engineers, and includes managerial tasks where models choose between technical implementation proposals.
Updated Aug 11, 2026
Higher score ranks better on this benchmark.
| 01 | OP | 100.0% | 100.0% | 6 | C | |
| 02 | OP | 81.4% | 80.0% | 6 | C | |
| 03 | OP | 74.6% | 60.0% | 6 | C | |
| 04 | OP | 17.4% | 40.0% | 6 | C | |
| 05 | OP | 12.4% | 20.0% | 6 | C | |
| 06 | OP | 7.4% | 0.0% | 6 | C |
A closer view of the leading scores on this benchmark.
The leading models and scores on this benchmark.
What SWE-Lancer (IC-Diamond subset) measures and how its scores work.
SWE-Lancer (IC-Diamond subset) is a benchmark of real-world freelance software engineering tasks from Upwork, ranging from $50 bug fixes to $32,000 feature implementations. It evaluates AI models on independent engineering tasks using end-to-end tests triple-verified by experienced software engineers, and includes managerial tasks where models choose between technical implementation proposals.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
Common questions about SWE-Lancer (IC-Diamond subset).
GPT-5 is currently ranked first with 100.0%.
SWE-Lancer (IC-Diamond subset) is a benchmark of real-world freelance software engineering tasks from Upwork, ranging from $50 bug fixes to $32,000 feature implementations. It evaluates AI models on independent engineering tasks using end-to-end tests triple-verified by experienced software engineers, and includes managerial tasks where models choose between technical implementation proposals.
Yes. Higher values rank better for this benchmark.
6 model results are currently shown.
Yes. This benchmark can contribute to the current LLMBoard capability score.