AgentWebBench (Centralized Retrieval) - Web Search: leaderboard
AgentWebBench (2026) models the Agentic Web: a user agent serves information needs over 100 ClueWeb22-B websites by coordinating with content agents or tools that each cover one website, with every LLM in thinking mode. This board is NDCG@5 in percent on MS MARCO web search queries whose relevant documents lie in the corpus, under centralized retrieval over one index of all websites, the reference without access constraints; higher is better.
Metric: NDCG@5 (%) on 354 MS MARCO Web Search queries whose relevant documents lie in the 100-website ClueWeb22-B corpus, LLM user agent under centralized retrieval (the agent retrieves from one index over all 100 websites, the no-access-constraint reference), thinking mode on; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 7 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 53.77 | #152 | |
| 2 | GPT-5 Mini | 52.09 | #364 | |
| 3 | Qwen 3 30B A3B 2507 (Thinking) | 49.9 | #764 | |
| 4 | DeepSeek V3.2 | Thinking | 49.37 | #419 |
| 5 | Qwen 3 Next 80B A3B (Thinking) | 48.78 | #594 | |
| 6 | Qwen 3 4B 2507 (Thinking) | 47.65 | #1433 | |
| 7 | Qwen 3 14B | 47.59 | #1269 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=agentwebbench-centralized-retrieval-web-search · How It Works · Data refreshed daily, snapshot 2026-11-21.