GPT-5.6 Terra security benchmark rundown
GPT-5.6 Terra (OpenAI) ranks 14th of 26 on BlueBench with 69.9 overall at $1.01 per task. Track scores, value against peers, strengths, weaknesses, and when to use it.
Verdict
GPT-5.6 Terra ranks 14th of 26 on BlueBench with an overall score of 69.9, in the lower half of the leaderboard and 13.3 points behind Claude Opus 5.5. At $1.01 per task it is the 10th cheapest of 26, and DeepSeek V4.1 Flash scores higher (71.5) for less ($0.17). Against the 10 models priced within 2x of its cost per task, it scores within 3 points of the peer median overall; it is strongest against them in threat hunting. Malware Analysis, weighted 10%, is a secondary weak spot (−7.4 vs the peer median). Its results are among the most consistent across the benchmarks (reliability 45.6, 8th of 26).
- BlueBench score
- 69.9
- #14 of 26 models
- Cost per task
- $1.01
- #10 cheapest of 26
- Reliability
- 45.6
- #8 of 26 models
- Overall vs cost peers
- +2.9
- Off the cost–score frontier
Capability by track
Incident response, threat hunting, and detection engineering each carry 30% of the BlueBench score; malware analysis carries 10%.
GPT-5.6 Terra scores 69.9 overall, 14th of 26 on the capability leaderboard. Its best rank is in threat hunting (62.1, 12th of 26) and its worst is in malware analysis (73.1, 17th of 26).
| Track | Weight | Score | Rank | Peer median | vs peers |
|---|---|---|---|---|---|
| Incident Response | 30% | 69.1 | #14 of 26 | 64.2 | +4.9 |
| Threat Hunting | 30% | 62.1 | #12 of 26 | 55.0 | +7.1 |
| Detection Engineering | 30% | 77.4 | #13 of 26 | 77.0 | +0.4 |
| Malware Analysis | 10% | 73.1 | #17 of 26 | 80.5 | −7.4 |
Cost, reliability, and value
Cost is USD per evaluated task including repeated attempts. Reliability is 100 minus the mean score range across the six benchmarks.
GPT-5.6 Terra averages $1.01 per task, including repeated attempts: the 10th cheapest of 26 models with a cost figure. It is off the cost–score frontier: DeepSeek V4.1 Flash scores 71.5 at $0.17, and 2 other models are also both cheaper and at least as good. Its peer group is the 10 models priced within 2x of its cost per task. Their median overall score is 67.0, so GPT-5.6 Terra is in line with its price range (+2.9). Reliability, 100 minus its mean score range across the benchmarks, is 45.6 (8th of 26).
Cost peers: the 10 models priced within 2x of its cost per task
Strengths and weaknesses
Each track compared with the median of the 10 models priced within 2x of its cost per task. A gap of 5 points or more counts.
Against the 10 models priced within 2x of its cost per task, GPT-5.6 Terra is strongest in threat hunting (+7.1). It lags them in malware analysis (−7.4). Differences are versus the peer median score in each track; ±5 points is the threshold.
Beats its cost peers
- Threat Hunting62.1 (+7.1)
Lags its cost peers
- Malware Analysis73.1 (−7.4)
Nearest neighbors on the leaderboard
The models ranked just above and below it overall, with the largest track gaps.
- GLM-5.3 Flash sits 2 places above, 0.7 points higher overall (70.6), at 11% of the cost per task ($0.11); no track differs by 5 points or more.
- GPT-6.1 Terra sits one place above, 0.6 points higher overall (70.5), at 16% of the cost per task ($0.16); it is 9.1 points ahead in incident response.
- Claude Haiku 5.5 sits one place below, 1.4 points lower overall (68.5), at 59% of the cost per task ($0.59); it is 5.3 points behind in malware analysis.
- Kimi K3 sits 2 places below, 2.6 points lower overall (67.3), at 2.6x the cost per task ($2.62); it is 15.4 points behind in detection engineering.
| Rank | Model | Score | $/task |
|---|---|---|---|
| 12 | GLM-5.3 Flash | 70.6 | $0.11 |
| 13 | GPT-6.1 Terra | 70.5 | $0.16 |
| 14 | GPT-5.6 Terra | 69.9 | $1.01 |
| 15 | Claude Haiku 5.5 | 68.5 | $0.59 |
| 16 | Kimi K3 | 67.3 | $2.62 |
Best for
Alternatives are limited to models costing at most 1.5x as much, except the overall leader and the cheaper-model pick.
- Use itUse it for threat hunting (62.1, +7.1) at its price: that track beats the median of the 10 models priced within 2x of its cost per task by 5 points or more.
- InsteadUse DeepSeek V4.1 Flash instead in most cases: it scores 71.5 overall (+1.6) at $0.17 per task (17% of the cost per task).
- InsteadUse Claude Sonnet 5.5 instead if malware analysis is the priority: it scores 87.8 there (vs 73.1) at $1.01 per task.
- InsteadUse Claude Opus 5.5 instead if the highest overall score matters more than cost: it scores 83.2 (13.3 points higher) at $3.54 per task.
Notable benchmark results
Each BlueBench benchmark is a separate intrusion, real or simulated, with its own tasks and grading.
- 3rd place on Intrusion-001 (macOS Threat Investigation): 97.5 (3rd of 26).
- Its scores range from 44.8 on Simulation-001 to 97.5 on Intrusion-001, a spread of 52.8 points.
| Benchmark | Score | Rank |
|---|---|---|
| Simulation-003Impact & Exfiltration | 66.3 | #13 of 26 |
| Simulation-002Identity & Active Directory Attacks | 65.7 | #13 of 26 |
| Simulation-001Initial Access & Command-and-Control | 44.8 | #21 of 26 |
| Intrusion-003AWS Cloud Intrusion | 82.9 | #8 of 26 |
| Intrusion-002Windows Enterprise Intrusion | 68.6 | #15 of 26 |
| Intrusion-001macOS Threat Investigation | 97.5 | #3 of 26 |
About this rundown
- Provider
- OpenAI · closed weight
- Tested
- Sep 11, 2026
- Suite
- 74 tasks across 6 benchmarks, evaluated in a common harness
- Data updated
- Oct 27, 2026
- How it is written
- Generated from the BlueBench result files with fixed rules: cost peers are models priced within 2x per task (or the nearest by cost when fewer than three are), a track is a strength or weakness at 5+ points from the peer median, and ranks share a place on ties. The text is regenerated whenever the results change.
AI for the blue team.
Run Cotool's harness in your environment to get real security work done