K-BrowseComp: A Web Browsing Agent Benchmark Grounded in Korean Contexts
K-BrowseComp is a new 400-problem benchmark for evaluating web-browsing agents in Korean-language contexts, with a 300-problem manually validated subset and a 100-problem adversarially constructed synthetic split. Frontier models including GPT-5.5, DeepSeek-V4-Pro, and GLM-5.1 achieve only 30–46% on the verified subset, a significant drop from English BrowseComp performance, while Korean proprietary models score 0–10%. The benchmark exploits the asymmetry between problem creation and solving difficulty, and the adversarial synthetic split caps the strongest model at 26%, positioning it as a targeted stress test for agentic web-browsing capability.
Related guides (3)
Related events (8)
BrowseComp: a benchmark for browsing agents
OpenAI has released BrowseComp, a benchmark designed to evaluate the capabilities of web-browsing AI agents. The benchmark appears to target the ability of agents to navigate and retrieve information from the web. As a Tier 1 source announcement, this represents OpenAI's effort to establish evaluation standards for agentic browsing behavior. Details on task structure, difficulty, and baseline results are not provided in the body text.
DeepWeb-Bench: A Hard Deep Research Benchmark Requiring Cross-Source Evidence and Long-Horizon Derivation
DeepWeb-Bench is a new benchmark designed to stress-test frontier language models on deep research tasks—open-web search, evidence collection, and multi-step derivation—where existing benchmarks have become saturated. The benchmark evaluates nine frontier models across four capability families (Retrieval, Derivation, Reasoning, Calibration) and finds that retrieval is not the primary bottleneck; derivation and calibration failures account for over 70% of errors. Strong models fail via incomplete derivation while weak models fail via hallucinated precision, and models show genuine domain specialization with low cross-model agreement (rho = 0.61). The benchmark, rubrics, and evaluation code are publicly released.
DeepSWE, ProgramBench, and ITBench-AA emerge as harder successors to SWE-bench for agent evaluation
Three new benchmarks — DeepSWE (by Datacurve), ProgramBench (Meta/Stanford/Harvard), and ITBench-AA (IBM/Artificial Analysis) — are positioned as more rigorous replacements for the SWE-bench family, which models have largely saturated. DeepSWE tests feature implementation using private codebases and human-written problems; ProgramBench evaluates agents' ability to recreate functional programs from scratch; ITBench-AA measures root-cause diagnosis in real-world IT incident scenarios. Current top performers include GPT-5.5 (70% on DeepSWE), Claude Opus 4.7 (46.7% on ITBench-AA), and Claude Opus 4.7 (3% on ProgramBench at the 95% pass threshold), illustrating that even frontier models have substantial headroom.
T1-Bench: Multi-scenario agent benchmark across 25 real-world domains
T1-Bench is a new benchmark for evaluating agentic LLM systems in realistic customer-facing, multi-domain environments, covering 25 domains of varying difficulty with interleaved multi-turn scenarios. The authors evaluate 12 proprietary and open-weight models and combine automatic evaluation with human judgments. The benchmark targets gaps in existing agent evals around task complexity, domain diversity, and compositional reasoning across multi-step interactions.
SearchGEO framework measures LLM search agent vulnerability to web content manipulation
Researchers introduce SearchGEO, a controlled evaluation framework for measuring endorsement corruption in LLM-based web-search agents, combining a manipulation pipeline, five-mode attack taxonomy, and multiple output metrics. Evaluating 13 LLM backends on 308 cases each, they find attack success rates ranging from 0.0% on Claude-Sonnet-4.6 to 31.4% on Gemini-3-Flash, with model-family-specific vulnerability patterns. An auxiliary probe escalating endorsement to install commands reveals a behavioral split: Claude over-rejects while GPT over-trusts. The findings argue for treating adversarial search content robustness as a first-class safety evaluation dimension for deployed agents.
AARRI-Bench evaluates frontier LLMs and agents on granular research-intern-level tasks
Researchers introduce AARR (Act As a Real Researcher), a new benchmark series targeting whether AI agents can emulate the professionalism, thoroughness, and nuanced judgment of human researchers in granular research scenarios—not just macro-level task execution. The first benchmark, AARRI-Bench, tests frontier models and agentic harnesses, finding that even the best configuration (Mini-SWE-Agent with Claude Opus 4.7) achieves only 68.3% success, frequently missing subtle but critical details obvious to human researchers. The work argues that closing the gap requires deeper modeling of research behavior rather than more complex scaffolding.
Benchmark Agent: Autonomous system for end-to-end benchmark construction
Researchers introduce Benchmark Agent, a fully autonomous agentic system that orchestrates the complete benchmark construction pipeline — from query analysis and subtask design to data annotation and quality control. The system was used to produce 15 benchmarks spanning text understanding, multimodal understanding, and domain-specific reasoning, with evaluation via human judges, LLM-as-a-judge, and consistency checks. The work addresses two persistent problems in the field: the labor intensity of benchmark creation and rapid performance saturation after release. Code and a demo will be publicly released.
SWE-Explore: New benchmark isolates repository exploration capability in coding agents
SWE-Explore is a new benchmark targeting repository exploration as a distinct, fine-grained capability of coding agents, separate from end-to-end task resolution. It covers 848 issues across 10 programming languages and 203 open-source repositories, with line-level ground truth derived from successful agent trajectories. Evaluation across retrieval methods, coding agents, and specialized localizers finds that agentic explorers outperform classical retrieval, and that line-level coverage and efficient ranking remain the key differentiators at the frontier. The benchmark addresses a gap in SWE-bench-style evaluations that treat task resolution as a binary outcome.