| Benchmark | Model | Rank | Results |
|---|---|---|---|
| computer-use-agents-on-online-mind2web | Orchard-GUI-4B (SFT + RL) | #5 | Success Rate: 67.0 |
| computer-use-agents-on-webvoyager | Orchard-GUI-4B (SFT + RL) | #8 | Success Rate: 74.1 |
| code-generation-on-swe-bench-verified | Qwen3-30B-A3B-Thinking (Orchard-SWE, SFT+RL) (mini-swe-agent) | #51 | Accuracy: 67.5 |