BrowseComp-Plus (BrowseComp+)
We show this table for reference; we do not rank on it.
Evaluates research agents against a fixed document corpus, separately from live-web BrowseComp.
Benchmark score on BrowseComp+ — October 29, 2026
We compile the BrowseComp+ rows from provider self-reports. Mercury 3 is the only currently displayable sourced row, at 55.18%. We do not use these results to rank models overall.
1 modelAgenticCurrentDisplay onlyUpdated October 29, 2026
Benchmark score table (1 model)
| Rank | Model / configuration | Score | Parameters (B) | Open / closed |
|---|---|---|---|---|
| 1 | Mercury 3Inception | 55.18% | Not reported | Closed |
How to read this leaderboard
Operator receipt: 1 sourced row is currently displayable on this page; the only currently displayable sourced row is Mercury 3 at 55.18%.
Honest limit: Retrieval setup and agent budget can change results. Display-only and excluded from weighted ranking.
About BrowseComp+
Year
2025
Inception reports Mercury 3 at 55.18 with medium reasoning. Its launch chart does not identify the corpus revision, retrieval configuration, agent harness, tool budget, or trial count. The provider result stays separate from original BrowseComp and benchmark-owner configurations.
Freshness and provenance
Version
BrowseComp+ 2025
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does BrowseComp+ measure?
Evaluates research agents against a fixed document corpus, separately from live-web BrowseComp.
Which model scores highest on BrowseComp+?
Mercury 3 by Inception currently leads with a score of 55.18% on BrowseComp+.
How many models are evaluated on BrowseComp+?
1 AI models have published results on BrowseComp+ in the BenchLM catalog.
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 6,100+ subscribers.
One email each week. Unsubscribe anytime.