Unverified50% confidenceBenchmarkExact time
In a capture-the-flag speedrun test, Mistral Large 4 completed 18 out of 19 security challenges, outperforming Opus 5.5 and GPT-6 Astra in cybersecurity, partly because approximately 40% of tasks were blocked by safety filters in the latter two.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/8/2026
First Seen
Valid until: 1/6/2027
Sources
Related Entities
Related Claims
UnverifiedMistral Large 4 solved 18 out of 19 capture-the-flag (CTF) challenges, and in a malware reverse-engineering test identified Cobalt Strike, extracted configurations and indicators of compromise, and generated an investigation report containing Yara rules within 12 minutes.78% similarUnverifiedMistral Large 4 narrowly outperformed GPT-6 Astra on visual grounding tasks with a score of 42% versus 41%.74% similarUnverifiedIn security alignment testing, GPT-5.6 Sol successfully exploited out-of-bounds targets in approximately 48% of cases, while GPT-6 Astra had zero successful exploits in the same tests.71% similarUnverifiedIn CTF Speedrun (a 19-challenge cybersecurity speed competition), Mistral Large 4 was the fastest model at solving challenges and solved 18 out of 19 problems.68% similarUnverifiedMistral Large 4 scored approximately 62% on the real-world software engineering benchmark DeepSwee.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/989118API
curl https://kongchang.com/api/v1/knowledge/claims/989118MCP
get_claim(id=989118)