Skip to main content

GPT-6.1 Terra on BlueBench

GPT-6.1 Terra ranks 13th of 26 models on BlueBench, Cotool's benchmark for defensive security work, with an overall score of 70.5. It ranks highest in incident response (78.2, 8th of 26) and lowest in detection engineering (72.5, 17th of 26). It averages $0.16 per task, the 2nd lowest cost of the 26 models tested.

OpenAI · Closed weight · Updated Oct 2026
Methodology
BlueBench score
70.5
#13 of 26 models
Cost per task
$0.16
#2 cheapest of 26
Reliability
45.1
#9 of 26 models
Best-ranked track
#8
Incident Response · 78.2
Read the GPT-6.1 Terra 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 Response78.2#8 of 26GPT-6 Astra · 88.0
Threat Hunting59.6#15 of 26Claude Opus 5.5 · 76.1
Detection Engineering72.5#17 of 26GPT-6 Astra · 87.3
Malware Analysis74.1#15 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-6.1 Terra on the overall leaderboard.

ModelRankScoreCost / task
#1170.9$3.27
#1270.6$0.11
GPT-6.1 Terra
#1370.5$0.16
#1469.9$1.01
#1568.5$0.59

Frequently asked questions

How good is GPT-6.1 Terra at security operations?
GPT-6.1 Terra scores 70.5 overall on BlueBench and ranks 13th 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-6.1 Terra best at?
BlueBench track scores for GPT-6.1 Terra, best rank first. Incident Response: 78.2 (8th of 26); Malware Analysis: 74.1 (15th of 26); Threat Hunting: 59.6 (15th of 26); Detection Engineering: 72.5 (17th of 26).
How much does GPT-6.1 Terra cost per security task?
GPT-6.1 Terra averages $0.16 per BlueBench task, including repeated attempts. That is the 2nd lowest cost of 26 models. The best-scoring cheaper model is GLM-5.3 Flash at $0.11 per task with an overall score of 70.6.
How was GPT-6.1 Terra 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