Benchmark
BrowseComp Long Context 256k Benchmark
What is BrowseComp Long Context 256k?
BrowseComp is a benchmark for measuring the ability of agents to browse the web, comprising 1,266 questions that require persistently navigating the internet…
- Modality
- text
- Evaluates
- reasoning, search
- Openness
- unknown
- Importer
- LLM Stats
- Reported scores
- 2
Source provenance
BrowseComp Long Context 256k results and reported scores
LLM Stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| GPT-5 | OpenAI | 0.888 | 2025-08-07 | evidence |
| GPT-5.2 | OpenAI | 0.898 | 2025-12-11 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.