Benchmark Almanac

CyScenarioBench

Gemini 4 Argon 100% leads since October 2026
In use

Scenario-based long-horizon security challenges in sandboxes, organized in task, scenario and campaign tiers. Evaluation set kept private.

Scenario based evaluation of long-horizon offensive operations, scored stage by stage. Difficulty is rated with Irregular's SOLVE+ scoring system, which grades reconnaissance, OpSec, social engineering and planning against human skill levels. Irregular reports counts of 11 and 13 scenarios in different 2026 assessments, so the set grows. Not public.

Made by
Irregular
Released
December 2025
Kind
Capture-the-flag challenges
Level
Expert
Scored
Checked automatically
Answers
Agent task
Tasks
Kept private
Measures
Capability
Language
English
In and out
Text, Code → Code, Actions
Interaction
Agentic
Environment
Terminal, A simulator
Run by
A host
Strongest model tested
GPT-6.2 Astra #1 on the capability index
Last in a launch post
November 2026
Last model added
November 2026
18
Models scored in lab cards
1
Launch post lists it, from 1 lab
19
System cards report it, from 5 labs

Gemini 4 Argon leads with 100%

Best score per model as labs report them in their own cards, drawn striped. Labs pick their own settings, so these compare less cleanly than independent runs.

  1. 1 Gemini 4 Argon 100%
  2. 2 Claude Fable 5.2 76.3%
  3. 3 Claude Opus 5.5 67.6%
  4. 4 GPT-6.2 Astra 63%
  5. 5 GPT-6 Cyber 61%
  6. 6 GPT-6 Astra 59%
  7. 7 Claude Sonnet 5.5 46.1%
  8. 8 Claude Opus 5 33.7%
  9. 9 GPT-5.6 Sol 28%
  10. 10 GPT-5.5 26%
  11. 11 GPT-5.4 11%
  12. 12 Claude Haiku 5.5 3.30%
  13. 13 Muse Spark 1.3 2.00%
  14. 14 Muse Spark 1.1 0.50%
  15. 15 GPT-5.2 Codex 0%
  16. 16 GPT-5.3 Codex 0%
  17. 17 Muse Spark 0%
  18. 18 Beam 1.0 0%

Source: Labs’ model cards

#ModelScoreSettingDateSource
1 Gemini 4 Argon 100% Oct 2026 Lab model card
2 Claude Fable 5.2 76.3% Nov 2026 Lab model card
3 Claude Opus 5.5 67.6% Sep 2026 Lab model card
4 GPT-6.2 Astra 63% Nov 2026 Lab model card
5 GPT-6 Cyber 61% Nov 2026 Lab model card
6 GPT-6 Astra 59% Sep 2026 Lab model card
7 Claude Sonnet 5.5 46.1% Sep 2026 Lab model card
8 Claude Opus 5 33.7% Jul 2026 Lab model card
9 GPT-5.6 Sol 28% Jun 2026 Lab model card
10 GPT-5.5 26% Apr 2026 Lab model card
11 GPT-5.4 11% Mar 2026 Lab model card
12 Claude Haiku 5.5 3.30% Oct 2026 Lab model card
13 Muse Spark 1.3 2.00% Oct 2026 Lab model card
14 Muse Spark 1.1 0.50% Jul 2026 Lab model card
15 GPT-5.2 Codex 0% Dec 2025 Lab model card
16 GPT-5.3 Codex 0% Feb 2026 Lab model card
17 Muse Spark 0% Jun 2026 Lab model card
18 Beam 1.0 0% Oct 2026 Lab model card

8 models have held number one in labs’ own numbers since its release in December 2025

Lab claims count from the day the card came out, independent runs from the model’s release.

OpenAI Anthropic Google DeepMind
Lab claims Passed 85% in Oct 2026, by Gemini 4 Argon

Source: Labs’ model cards

Google DeepMind’s own number for Gemini 4 Argon passed 85% in October 2026

That was 10 months after its release. By the same age, 19% of the benchmarks from 2025 had passed 85%. All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
Reported by the lab in its model card Best so far
0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Gemini 4 Argon Claude Opus 5.5 GPT-6 Astra Claude Opus 5 GPT-5.6 Sol GPT-5.4 0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Gemini 4 Argon Claude Opus 5.5

Source: Labs’ model cards

One lab put it in a launch post in the last six months

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.

OpenAI
DateLaunch postScoreAs the post writes it
Nov 5, 2026 GPT-6 Cyber 61 CyScenarioBench, average success rate across 11 long-horizon challenges

19 system cards report it, from 5 labs

Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.

DateDocumentScoreAs the document writes itSection
Nov 17, 2026 GPT-6.2 Astra system card 63 Irregular CyScenarioBench, success rate, latest suite version (9 of 11 solved at least once) (2.1.2.2.3) Cyber risk
Nov 9, 2026 Claude Fable 5.2 system card 76.3 CyScenarioBench, Claude Mythos 5.2 completed 76.3% of all challenges (pre-release snapshot) Cyber risk
Nov 5, 2026 GPT-6 Cyber system card 61 CyScenarioBench, average success rate across 11 long-horizon challenges Cyber risk
Oct 23, 2026 Muse Spark 1.3 system card 2 CyScenarioBench pass@1, all 10 scenarios, Muse Spark 1.3 (max) (Table 1, Table 4). pass@20 = 30.0, 3 of 10 solved Cyber risk
Oct 13, 2026 Gemini 4 Argon system card 100 Irregular CyScenarioBench, challenges solved end-to-end 12/12; segments passed 77/82 (93.9%) (Figure 3.2, 3.2.2.3) Cyber risk
Oct 7, 2026 Claude Haiku 5.5 system card 3.3 CyScenarioBench, overall solve rate on the 10-challenge subset, pre-release snapshot, cyber mitigations off (3.3.2) Cyber risk
Oct 5, 2026 Beam 1.0 technical report 0 Irregular CTF suite, CyScenarioBench group, 0 of 10 solved Cyber risk
Sep 28, 2026 Claude Sonnet 5.5 system card 46.1 CyScenarioBench, average solve rate over a 10-challenge subset, cyber mitigations off (3.2.2, Figure 3.2.2.A) Cyber risk
Sep 22, 2026 Claude Opus 5.5 system card 67.6 CyScenarioBench, overall solve rate averaged over a 10-challenge subset, cyber mitigations off (3.3.2, Figure 3.3.2.A) Cyber risk
Sep 3, 2026 GPT-6 Astra system card 59 Irregular CyScenarioBench, average success rate (9 of 10 solved at least once) (10.1.2.6) Cyber risk
Jul 24, 2026 Claude Opus 5 system card 33.7 CyScenarioBench, overall solve rate averaged over a 9-challenge subset, mitigations off (3.3.4) Cyber risk
Jul 9, 2026 GPT-5.6 Sol system card 28 Irregular CyScenarioBench, average success, 7 of 11 challenges solved (9.1.2.5) Cyber risk
Jul 9, 2026 Muse Spark 1.1 system card 0.5 CyScenarioBench pass@1 (Table 1, Figure 11). pass@20 = 10.0, 1 of 10 scenarios Cyber risk
Jun 25, 2026 GPT-5.6 Sol system card 28 Irregular CyScenarioBench, average success, 7 of 11 challenges solved (9.1.2.5) Cyber risk
Jun 12, 2026 Muse Spark system card 0 CyScenarioBench, 10 scenarios (Table 1) Cyber risk
Apr 23, 2026 GPT-5.5 system card 26 Irregular CyScenarioBench, average success rate, 7 of 11 challenges solved (9.1.2.5) Cyber risk
Mar 5, 2026 GPT-5.4 system card 11 Irregular CyScenarioBench, average success rate, 5 of 11 challenges solved (5.1.2.4) Cyber risk
Feb 5, 2026 GPT-5.3 Codex system card 0 CyScenarioBench by Irregular, text says no challenges solved (5.1.2.4) Cyber risk
Dec 18, 2025 GPT-5.2 Codex system card addendum 0 Irregular CyScenarioBench, text says no challenges solved (5.1.2.4) Cyber risk

More in Cybersecurity

All 138 in Cybersecurity
  • ExploitGym In use 869 real-vulnerability exploit-development tasks (userspace, V8, Linux kernel). Success means capturing a flag through the intended vulnerability. GPT-6 Astra 42.4%
  • ExploitBench In use 41 recent V8 JavaScript-engine N-day vulnerabilities. Progress measured by 16 grader-verified capability flags across five tiers. GPT-6 Astra 100%
  • ExploitBench - Internal Port In use Agents develop working exploits for recently disclosed vulnerabilities to limit contamination. Scored as arbitrary code execution success rate. GPT-6.2 Astra 53.5%
  • Atomic Challenges In use Agents solve 22 skill-level offensive security challenges in vulnerability research, network attack simulation and evasion. Scored as challenges solved. GPT-6 Astra 90.9%
  • MiMo Cyber Bench In use Agents do in-house cybersecurity tasks that add to public vulnerability reproduction and exploitation benchmarks. The report gives no task count or grading details. MiMo-V2.6 Pro 81.7%
  • PatchBench (OpenAI) In use Models patch security flaws. Automated tests and model-based review check that each patch fixes the flaw, keeps the intended behavior and is ready to merge. GPT-6.2 Astra 79.7%

Data audit

6 of 7 checks pass, 1 warning. Look first at the Importer check. The record links a board on irregular.com that we don’t read yet. Every benchmark

Catalog record 2 pass
  • Pass Task access The tasks are private.
  • Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
  • Note Independent runs Only labs’ own numbers, so no one else has checked them. Scores from launch posts (1 score) and system cards (19 scores).
  • Pass Models scored 18 models we track.
  • Pass Newest score The newest is from November 2026, for GPT-6.2 Astra.
  • Pass Frontier models 6 of the top 10 have a score, led by GPT-6.2 Astra (#1).
  • Note Headroom Saturated. The best is 100%, past 85% since October 2026.
Leaderboard import 1 warning
  • Warning Importer The record links a board on irregular.com that we don’t read yet.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check