List your company

Question Answering on GAIA (Pass@4, Usage Metrics)

51Accuracy (Pass@4)SynPlanResearch-R1
5.333617.189329.04540.9007Jan 21, 2026Feb 10, 2026Mar 2, 2026Mar 22, 2026Apr 11, 2026May 1, 2026May 22, 2026

Evaluation results

MethodLinks
SynPlanResearch-R1Model Backbone=Qwen3-8B2026.0351—————
SearchGym-RL (Ours)Source=local, Backbone=Qwen-3-8B2026.0150.495,200.555.131.810—
SimpleDeepSearcherModel Backbone=Qwen3-8B2026.0345.2—————
SynPlanResearch-R1Model Backbone=Qwen3-4B2026.0343.9—————
SearchGym-RL (Ours-base)Source=local, Backbone=Qwen-2.5-7B-Base2026.0142.721,114.483.710.90—
Rejection SamplingModel Backbone=Qwen3-8B2026.0341.7—————
SimpleDeepSearcherModel Backbone=Qwen3-4B2026.0341.5—————
Search-R1Model Backbone=Qwen3-8B2026.0340.5—————
ASearcher-base-webSource=web, Backbone=Qwen-2.5-7B-Base2026.0138.83692.285.920.07——
Search-R1Model Backbone=Qwen3-4B2026.0336.2—————
Rejection SamplingModel Backbone=Qwen3-4B2026.0331.2—————
ASearcher-base-localSource=local, Backbone=Qwen-2.5-7B-Base2026.0126.211,022.387.030.030—
Search-o1Model Backbone=Qwen3-8B2026.0325.7—————
Search-o1Model Backbone=Qwen3-4B2026.0325—————
EPC-AWBackbone=Qwen3-Coder-30B2026.0515.22—————
RAGModel Backbone=Qwen3-4B2026.0315—————
Direct InferenceModel Backbone=Qwen3-8B2026.0314.6—————
RAGModel Backbone=Qwen3-8B2026.0313—————
Direct InferenceModel Backbone=Qwen3-4B2026.0311.9—————
RollbackBackbone=Qwen3-Coder-30B2026.0510.76————4.46
RetryBackbone=Qwen3-Coder-30B2026.059.17————6.05
No-Repair (AgentFlow)Backbone=Qwen3-Coder-30B2026.057.09————8.13