HippoCamp (Autonomous Agents): leaderboard

HippoCamp (2026) tests agents that search and reason over device-scale personal file systems built from three real user profiles, answering evidence-grounded questions judged by GPT-4o against reference answers. This board is answer accuracy in percent over all questions for autonomous agents that explore the file system with terminal tools; higher is better.

Metric: LLM-judge answer accuracy (%) on HippoCamp, all 581 queries (60 profiling and 521 factual-retention questions, sample-weighted), over three personal file-system profiles (Bei, Adam, Victoria; 42.4 GB, about 2K files), each answer judged correct or not by GPT-4o against the reference answer; autonomous-agent setting: Qwen3-VL-8B-Instruct, Gemini 2.5 Flash and GPT-5.2 drive the benchmark vacuum-Docker terminal interface, and ChatGPT agent runs in its hosted product mode; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 4 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1ChatGPT Agent55.9
2GPT-5.244.1#185
3Gemini 2.5 Flash22#463
4Qwen 3 VL 8B Instruct11.7#1029

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=hippocamp-autonomous-agents · How It Works · Data refreshed daily, snapshot 2026-11-21.