| Paper | Code | Success Rate | ModelName | ReleaseDate |
|---|---|---|---|---|
| PANDO: Efficient Multimodal AI Agents via Online Skill Distillation | 58.3 | PANDO (Opus 4.6 + GPT-5.2) | 2026-05-26 | |
| CLIFT: Conformal Self-Verification for Web Agent Training and Test-Time Scaling | 53.7 | GPT-5.5 + CLIFT+CTS | 2026-10-05 | |
| WebChallenger: A Reliable and Efficient Generalist Web Agent | ✓ Link | 48.7 | WebChallenger (GLM-4-32B) | 2026-06-09 |
| Structured Distillation of Web Agent Capabilities Enables Generalization | ✓ Link | 33.9 | A3-Qwen3.5-9B | 2026-04-09 |