{"slug": "evaluating-gpt-5-6-terra-pro-on-hack-the-box-challenges", "title": "Evaluating GPT-5.6 Terra Pro on Hack The Box Challenges", "summary": "GPT-5.6 Terra Pro solved 14 of 16 Hack The Box challenges, achieving a benchmark score of 75.1% with a median cost of $0.91 per challenge, according to the HTB-Challenger Benchmark. The model solved all Very Easy, Easy, and Medium challenges but only 2 of 4 Hard challenges, with a 44.4% benchmark score on Hard difficulty.", "body_md": "# Evaluating GPT‑5.6 Terra Pro on Hack The Box Challenges\n\nFor some personal observations about this model, please see the\n\n[Solving Hack The Box Challenges with GPT-5.6]post.\n\nThis blog post is part of a series of tests for the\n\n[HTB-Challenger Benchmark]The HTB-Challenger Benchmark evaluates LLMs’ ability to find and exploit security vulnerabilities. It tests models against selected Hack The Box challenges of varying difficulty and measures their performance. For more information, visit the[HTB-Challenger Benchmark page]. . See the[benchmark results page]for all results and the[benchmark methodology]to learn how the benchmark is calculated.\n\n## Cost vs. Benchmark Score\n\nThe highlighted point is this model. Models closer to the upper-left achieve a higher benchmark score at a lower median cost per challenge.\n\n## Overall benchmark results\n\n**Number of challenges:** 16**Number of solved challenges:** 14**Number of false positives:** 0**Runs where the model gave up:** 2**Runs that reached the step or cost limit:** 0**Runs where the model got stuck:** 0**Benchmark score:** 75.1%\n\n| Metric | Per challenge (median) | Total |\n|---|---|---|\n| Model steps | 11 | 353 |\n| Model cost | $0.91 | $29.93 |\n| Duration | 00:08:22 | 02:30:03 |\n| Number of input tokens | 0.64M | 22.53M |\n| Number of output tokens | 0.03M | 0.83M |\nNumber of `read_file` tool calls |\n0.0 | 20 |\nNumber of `write_file` tool calls |\n0.0 | 18 |\nNumber of `execute_command` tool calls |\n13.5 | 351 |\nNumber of `web_search` tool calls |\n0.0 | 4 |\n\n## Results by challenge difficulty\n\nAll resource-usage metrics are medians per challenge.\n\n| Metric | Very Easy | Easy | Medium | Hard |\n|---|---|---|---|---|\nResults |\n||||\n| Number of challenges | 4 | 4 | 4 | 4 |\n| Number of solved challenges | 4 | 4 | 4 | 2 |\n| Number of false positives | 0 | 0 | 0 | 0 |\n| Runs where the model gave up | 0 | 0 | 0 | 2 |\n| Runs that reached the step or cost limit | 0 | 0 | 0 | 0 |\n| Runs where the model got stuck | 0 | 0 | 0 | 0 |\n| Benchmark score | 97.3% | 97.1% | 94.0% | 44.4% |\nMedian per challenge |\n||||\n| Model steps | 5 | 9.5 | 20.5 | 29.5 |\n| Model cost | $0.16 | $0.53 | $1.95 | $2.46 |\n| Duration | 00:01:10 | 00:03:35 | 00:09:53 | 00:18:01 |\n| Number of input tokens | 0.08M | 0.39M | 1.39M | 1.98M |\n| Number of output tokens | 0.01M | 0.02M | 0.06M | 0.07M |\nNumber of `read_file` tool calls |\n0.5 | 0.5 | 0.0 | 0.0 |\nNumber of `write_file` tool calls |\n0.0 | 0.5 | 0.0 | 0.0 |\nNumber of `execute_command` tool calls |\n4.0 | 9.0 | 19.5 | 31.0 |\nNumber of `web_search` tool calls |\n0.5 | 0.0 | 0.0 | 0.0 |", "url": "https://wpnews.pro/news/evaluating-gpt-5-6-terra-pro-on-hack-the-box-challenges", "canonical_source": "https://theaq.blog/2026/08/16/solving-htb-challenges-with-openai-gpt-5.6-terra-pro.html", "published_at": "2026-08-16 16:10:31+00:00", "updated_at": "2026-08-21 10:43:05.283237+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-research", "ai-tools"], "entities": ["GPT-5.6 Terra Pro", "Hack The Box", "HTB-Challenger Benchmark"], "alternates": {"html": "https://wpnews.pro/news/evaluating-gpt-5-6-terra-pro-on-hack-the-box-challenges", "markdown": "https://wpnews.pro/news/evaluating-gpt-5-6-terra-pro-on-hack-the-box-challenges.md", "text": "https://wpnews.pro/news/evaluating-gpt-5-6-terra-pro-on-hack-the-box-challenges.txt", "jsonld": "https://wpnews.pro/news/evaluating-gpt-5-6-terra-pro-on-hack-the-box-challenges.jsonld"}}