LongSeeker: Elastic Context Orchestration for Long-Horizon Search Agents (May 2026) logo

LongSeeker: Elastic Context Orchestration for Long-Horizon Search Agents (May 2026)

Free

Context-ReAct paradigm with five atomic operations (Skip, Compress, Rollback, Snippet, Delete) for adaptive context management; proves expressive completeness of Compress; LongSeeker achieves 61.5% on BrowseComp and 62.5% on BrowseComp-ZH, substantially outperforming Tongyi DeepResearch and AgentFol

FreeFree tier
Type
Open Source

About LongSeeker: Elastic Context Orchestration for Long-Horizon Search Agents (May 2026)

LongSeeker is a long-horizon search agent that introduces the Context-ReAct paradigm for adaptive context management. It provides five atomic operations (Skip, Compress, Rollback, Snippet, Delete) to dynamically reshape working context, preserving important evidence, summarizing resolved information, and discarding unhelpful branches. The Compress operator is proven to be expressively complete. LongSeeker is fine-tuned from Qwen3-30B-A3B on 10k synthesized trajectories and achieves state-of-the-art results: 61.5% on BrowseComp and 62.5% on BrowseComp-ZH, substantially outperforming existing agents like Tongyi DeepResearch and AgentFold.

Key Features

Context-ReAct paradigm integrating reasoning, context management, and tool use
Five atomic operations: Skip, Compress, Rollback, Snippet, Delete
Expressive completeness of the Compress operator
Fine-tuned from Qwen3-30B-A3B on 10k synthesized trajectories
Achieves 61.5% on BrowseComp and 62.5% on BrowseComp-ZH

Pros & Cons

Pros
  • State-of-the-art performance on BrowseComp benchmarks
  • Provably expressive context management
  • Reduces hallucination risk and generation cost via efficient operators
  • Open-source and accessible via arXiv
Cons
  • Only evaluated on synthetic trajectories and specific benchmarks
  • Requires large base model (30B parameters)
  • Not a ready-to-use product, requires implementation

Best For

Long-horizon search tasksMulti-step tool use in browsingAdaptive context management for agentsResearch in agentic search

FAQ

What is LongSeeker?
LongSeeker is a long-horizon search agent that uses the Context-ReAct paradigm to manage working context adaptively.
What atomic operations does Context-ReAct provide?
Context-ReAct provides five atomic operations: Skip, Compress, Rollback, Snippet, and Delete.
How does LongSeeker perform on benchmarks?
LongSeeker achieves 61.5% on BrowseComp and 62.5% on BrowseComp-ZH, outperforming Tongyi DeepResearch and AgentFold.