CTI-REALM
We show this table for reference; we do not rank on it.
A cybersecurity benchmark that measures whether an agent can turn raw threat-intelligence reports into working detection rules.
Benchmark score on CTI-REALM — November 13, 2026
We compile the CTI-REALM rows from provider self-reports. Fugu Cyber leads the table at 72.1%. We do not use these results to rank models overall.
1 modelAgenticCurrentDisplay onlyUpdated November 13, 2026
Benchmark score table (1 model)
The one result on this page was reported by the lab that built the model.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | Fugu CyberSakana AIReported by Sakana AI | 72.1% | Reported by Sakana AI | Not reported | Closed |
- Lab-reported
- Benchmark-run
- Independent
About CTI-REALM
Year
2026
Tasks
Threat-intelligence-to-detection-rule workflows
Format
Success rate
Difficulty
Professional cyber threat detection
Sakana AI describes CTI-REALM as a real-world security benchmark for translating cyber threat intelligence into executable detection rules. We store exact provider-reported results as display-only evidence until the benchmark owner publishes a stable, independently verifiable leaderboard artifact.
CTI-REALM is one of the cyber benchmarks we track. The cybersecurity benchmark comparison sets it beside the others, one table per security job.
Freshness and provenance
Version
CTI-REALM 2026
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 CTI-REALM measure?
A cybersecurity benchmark that measures whether an agent can turn raw threat-intelligence reports into working detection rules.
Which model scores highest on CTI-REALM?
Fugu Cyber by Sakana AI currently leads with a score of 72.1% on CTI-REALM.
How many models are evaluated on CTI-REALM?
1 AI models have published results on CTI-REALM 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 7,000+ subscribers.
One email each week. Unsubscribe anytime.