Skip to main content

GPT-6 Sol on BlueBench

GPT-6 Sol ranks 10th of 26 models on BlueBench, Cotool's benchmark for defensive security work, with an overall score of 71.1. It ranks highest in incident response (77.9, 9th of 26) and lowest in threat hunting (56.2, 16th of 26). It averages $1.74 per task, the 15th lowest cost of the 26 models tested.

OpenAI · Closed weight · Updated Oct 2026
Methodology
BlueBench score
71.1
#10 of 26 models
Cost per task
$1.74
#15 cheapest of 26
Reliability
46.3
#7 of 26 models
Best-ranked track
#9
Incident Response · 77.9
Read the GPT-6 Sol rundown →Verdict, strengths and weaknesses against similar-cost models, and when to pick something else.

Scores by track

BlueBench weights incident response, threat hunting, and detection engineering at 30% each and malware analysis at 10%.

TrackScoreRankTrack leader
Incident Response77.9#9 of 26GPT-6 Astra · 88.0
Threat Hunting56.2#16 of 26Claude Opus 5.5 · 76.1
Detection Engineering78.4#11 of 26GPT-6 Astra · 87.3
Malware Analysis73.8#16 of 26Claude Opus 5.5 · 89.4

Scores by benchmark

Each BlueBench benchmark is a separate intrusion, real or simulated, with its own tasks and grading.

Models ranked nearby

The models closest to GPT-6 Sol on the overall leaderboard.

ModelRankScoreCost / task
#874.9$1.56
#971.5$0.17
GPT-6 Sol
#1071.1$1.74
#1170.9$3.27
#1270.6$0.11

Frequently asked questions

How good is GPT-6 Sol at security operations?
GPT-6 Sol scores 71.1 overall on BlueBench and ranks 10th of 26 models on incident response, threat hunting, detection engineering, and malware analysis. The top model, Claude Opus 5.5, scores 83.2.
Which security tasks is GPT-6 Sol best at?
BlueBench track scores for GPT-6 Sol, best rank first. Incident Response: 77.9 (9th of 26); Detection Engineering: 78.4 (11th of 26); Malware Analysis: 73.8 (16th of 26); Threat Hunting: 56.2 (16th of 26).
How much does GPT-6 Sol cost per security task?
GPT-6 Sol averages $1.74 per BlueBench task, including repeated attempts. That is the 15th lowest cost of 26 models. The best-scoring cheaper model is Claude Sonnet 5.5 at $1.01 per task with an overall score of 77.5.
How was GPT-6 Sol tested?
Every model runs the same 74 tasks across 6 BlueBench benchmarks, built from real intrusion logs and simulated enterprise environments, using the same harness, prompts, tools, and number of trials. Reports are graded against hidden rubrics and detection rules are re-run against the evidence.

Full leaderboards

AI for the blue team.

Run Cotool's harness in your environment to get real security work done

Book a demo