Web Navigation Question Answering on WebWalker QA
76.5AccuracySMTL-30B-300
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| SMTL-30B-300Model Category=Trained Agents (≥30B)2026.03 | 76.5 | — | — | — | |
| Tongyi-DeepResearch 30BModel Access Type=Open-Source2026.02 | 72.2 | — | — | — | |
| Tongyi-DR-30BModel Category=Trained Agents (≥30B)2026.03 | 72.2 | — | — | — | |
| OpenAI-o3Model Category=Foundation Models with Tools2026.03 | 71.7 | — | — | — | |
| TodoEvolveBackbone=Kimi-K22026.02 | 70 | 0.0167 | 216.59 | 7.7 | |
| Marco-DR-8BModel Category=Trained Agents (≤8B)2026.03 | 69.6 | — | — | — | |
| AgentCPM-Explore-4BModel Access Type=Open-Source2026.02 | 68.1 | — | — | — | |
| AgentCPM-Explore-4BModel Category=Trained Agents (≤8B)2026.03 | 68.1 | — | — | — | |
| Merged-Model-4BModel Access Type=Open-Source2026.02 | 65.4 | — | — | — | |
| JoyAgentBackbone=Kimi-K22026.02 | 63.33 | 0.0028 | 212.83 | 4 | |
| WebExplorer-8B-RLModel Category=Trained Agents (≤8B)2026.03 | 62.7 | — | — | — | |
| MiroThinker-v1.0-72BModel Category=Trained Agents (≥30B)2026.03 | 62.1 | — | — | — | |
| Claude-4-SonnetModel Category=Foundation Models with Tools2026.03 | 61.7 | — | — | — | |
| MiroThinker-v1.0-30BModel Category=Trained Agents (≥30B)2026.03 | 61 | — | — | — | |
| MiroThinker 8BModel Access Type=Open-Source2026.02 | 60.6 | — | — | — | |
| MiroThinker-v1.0-8BModel Category=Trained Agents (≤8B)2026.03 | 60.6 | — | — | — | |
| Flash-SearcherBackbone=Kimi-K22026.02 | 60 | 0.0134 | 164.78 | 5.3 | |
| OWLBackbone=Kimi-K22026.02 | 53.33 | 0.0062 | 127.63 | 3.8 | |
| WebDancer-QwQ-32BModel Access Type=Open-Source2026.02 | 47.9 | — | — | — | |
| AgentOrchestraBackbone=Kimi-K22026.02 | 46.67 | 0.0112 | 137.69 | 6.5 | |
| OAgentsBackbone=Kimi-K22026.02 | 33.33 | 0.0236 | 150.74 | 7.2 | |
| FlowSearchBackbone=Kimi-K22026.02 | 30 | 0.0053 | 94.79 | 4 | |
| Co-SightBackbone=Kimi-K22026.02 | 16.67 | 0.0013 | 190.52 | 2.1 | |
| AgentOccamModel=Qwen3, Backbone size=4B2026.04 | — | — | — | 12.84 | |
| AgentOccamModel=Qwen3, Backbone size=8B2026.04 | — | — | — | 6.88 | |
| AgentOccamModel=Qwen3, Backbone size=14B2026.04 | — | — | — | 7.65 | |
| AgentOccamModel=Qwen3, Backbone size=32B2026.04 | — | — | — | 14.33 | |
| AgentOccamModel=GPT-5 mini2026.04 | — | — | — | 10.09 | |
| MANGOModel=Qwen3, Backbone size=4B2026.04 | — | — | — | 5.92 | |
| MANGOModel=Qwen3, Backbone size=8B2026.04 | — | — | — | 10.89 | |
| MANGOModel=Qwen3, Backbone size=14B2026.04 | — | — | — | 7.75 | |
| MANGOModel=Qwen3, Backbone size=32B2026.04 | — | — | — | 7.49 | |
| MANGOModel=GPT-5 mini2026.04 | — | — | — | 19.13 | |
| WebWalkerModel=Qwen3, Backbone size=4B2026.04 | — | — | — | 8.71 | |
| WebWalkerModel=Qwen3, Backbone size=8B2026.04 | — | — | — | 9.62 | |
| WebWalkerModel=Qwen3, Backbone size=14B2026.04 | — | — | — | 7.92 | |
| WebWalkerModel=Qwen3, Backbone size=32B2026.04 | — | — | — | 8.62 | |
| WebWalkerModel=GPT-5 mini2026.04 | — | — | — | 10.38 |