{"slug": "solving-hack-the-box-challenges-with-gpt-5-6-luna-pro", "title": "Solving Hack The Box Challenges with GPT-5.6 Luna Pro", "summary": "OpenAI's GPT-5.6 Luna Pro solved 11 of 16 Hack The Box challenges, scoring 55.0% on the HTB-Challenger Benchmark, compared with GPT-5.6 Luna which solved only one Medium and no Hard challenges. The model, served via OpenRouter.ai with reasoning.mode set to 'pro', is approximately five times more expensive than Luna but remains cheap, with a median cost of $0.10 per challenge and total cost of $3.05. It solved all Very Easy and Easy challenges, 2 of 4 Medium, and 1 of 4 Hard, with no false positives.", "body_md": "# Solving Hack The Box Challenges with GPT-5.6 Luna Pro\n\nWhat the heck is **GPT-5.6 Luna Pro**, I hear you asking. That’s a very good question! To answer it, let me quote its [description on OpenRouter.ai](https://openrouter.ai/openai/gpt-5.6-luna-pro): “* GPT-5.6 Luna Pro is the same underlying model as GPT-5.6 Luna, served with *”\n\n`reasoning.mode`\n\nset to `pro`\n\nfor higher-quality responses on complex tasks.Okay, so how much better, and how much more expensive, is it compared with **GPT-5.6 Luna**, I hear you asking now. And that’s exactly what I can tell you.\n\nThis blog post is part of a series of tests for the HTB-Challenger Benchmark. See the\n\n[benchmark results page]for all results and the[benchmark methodology]to learn how the benchmark is calculated.\n\nI tested **GPT-5.6 Luna** in my [previous post](/2026/08/10/solving-htb-challenges-with-openai-gpt-5.6-luna.html), so I have the numbers ready. In a nutshell:\n\n**Luna Pro** is approximately five times more expensive than**Luna**. It may sound dramatic, but** Luna**is such a cheap model that** Luna Pro**is still very cheap.- In this benchmark run,\n**Luna Pro** performed much better than**Luna**. If you look below at the results for the more difficult challenges, you can see that** Luna Pro**solved some Medium and Hard challenges, while** Luna**solved only one Medium challenge and no Hard challenges.\n\nI believe the main use case for **Luna Pro** is if you are tied to the OpenAI ecosystem and need a model for offensive-security or CTF-style tasks. In my tests, **GPT-5.6 Terra** and **GPT-5.6 Sol** refused to respond to most offensive-security requests, so **GPT-5.6 Luna Pro** was the most useful OpenAI option for this kind of work.\n\n## Overall benchmark results\n\n**Number of challenges:** 16**Number of solved challenges:** 11**Number of false positives:** 0**Runs where the model gave up:** 4**Runs that reached the step or cost limit:** 1**Runs where the model got stuck:** 0**Benchmark score:** 55.0%\n\n| Metric | Per challenge (median) | Total |\n|---|---|---|\n| Model steps | 21.5 | 554 |\n| Model cost | $0.10 | $3.05 |\n| Duration | 00:08:55 | 03:48:35 |\n| Number of input tokens | 1.50M | 42.93M |\n| Number of output tokens | 0.07M | 1.63M |\nNumber of `read_file` tool calls |\n4.0 | 186 |\nNumber of `write_file` tool calls |\n0.0 | 18 |\nNumber of `execute_command` tool calls |\n36.5 | 865 |\nNumber of `web_search` tool calls |\n0.0 | 3 |\n\n## Results by challenge difficulty\n\n### Very Easy challenges\n\n**Number of challenges:** 4**Number of solved challenges:** 4**Number of false positives:** 0**Runs where the model gave up:** 0**Runs that reached the step or cost limit:** 0**Runs where the model got stuck:** 0**Benchmark score:** 100.0%\n\n| Metric | Per challenge (median) | Total |\n|---|---|---|\n| Model steps | 6 | 38 |\n| Model cost | $0.01 | $0.14 |\n| Duration | 00:01:43 | 00:12:53 |\n| Number of input tokens | 0.16M | 1.89M |\n| Number of output tokens | 0.01M | 0.08M |\nNumber of `read_file` tool calls |\n0.5 | 7 |\nNumber of `write_file` tool calls |\n0.0 | 1 |\nNumber of `execute_command` tool calls |\n6.0 | 73 |\nNumber of `web_search` tool calls |\n0.0 | 0 |\n\n### Easy challenges\n\n**Number of challenges:** 4**Number of solved challenges:** 4**Number of false positives:** 0**Runs where the model gave up:** 0**Runs that reached the step or cost limit:** 0**Runs where the model got stuck:** 0**Benchmark score:** 100.0%\n\n| Metric | Per challenge (median) | Total |\n|---|---|---|\n| Model steps | 21.5 | 80 |\n| Model cost | $0.09 | $0.34 |\n| Duration | 00:08:46 | 00:37:25 |\n| Number of input tokens | 1.38M | 4.94M |\n| Number of output tokens | 0.06M | 0.22M |\nNumber of `read_file` tool calls |\n1.5 | 12 |\nNumber of `write_file` tool calls |\n0.0 | 0 |\nNumber of `execute_command` tool calls |\n26.5 | 99 |\nNumber of `web_search` tool calls |\n0.5 | 3 |\n\n### Medium challenges\n\n**Number of challenges:** 4**Number of solved challenges:** 2**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:** 50.0%\n\n| Metric | Per challenge (median) | Total |\n|---|---|---|\n| Model steps | 33 | 175 |\n| Model cost | $0.16 | $1.02 |\n| Duration | 00:13:37 | 01:10:21 |\n| Number of input tokens | 2.30M | 14.23M |\n| Number of output tokens | 0.10M | 0.54M |\nNumber of `read_file` tool calls |\n8.5 | 75 |\nNumber of `write_file` tool calls |\n0.5 | 3 |\nNumber of `execute_command` tool calls |\n47.5 | 265 |\nNumber of `web_search` tool calls |\n0.0 | 0 |\n\n### Hard challenges\n\n**Number of challenges:** 4**Number of solved challenges:** 1**Number of false positives:** 0**Runs where the model gave up:** 2**Runs that reached the step or cost limit:** 1**Runs where the model got stuck:** 0**Benchmark score:** 25.0%\n\n| Metric | Per challenge (median) | Total |\n|---|---|---|\n| Model steps | 73.5 | 261 |\n| Model cost | $0.45 | $1.55 |\n| Duration | 00:28:35 | 01:47:55 |\n| Number of input tokens | 6.36M | 21.87M |\n| Number of output tokens | 0.22M | 0.79M |\nNumber of `read_file` tool calls |\n16.5 | 92 |\nNumber of `write_file` tool calls |\n2.5 | 14 |\nNumber of `execute_command` tool calls |\n121.0 | 428 |\nNumber of `web_search` tool calls |\n0.0 | 0 |", "url": "https://wpnews.pro/news/solving-hack-the-box-challenges-with-gpt-5-6-luna-pro", "canonical_source": "https://theaq.blog/2026/08/10/solving-htb-challenges-with-openai-gpt-5.6-luna-pro.html", "published_at": "2026-08-10 16:51:46+00:00", "updated_at": "2026-08-12 21:46:20.476656+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-products", "ai-research"], "entities": ["OpenAI", "GPT-5.6 Luna Pro", "GPT-5.6 Luna", "OpenRouter.ai", "Hack The Box", "HTB-Challenger Benchmark", "GPT-5.6 Terra", "GPT-5.6 Sol"], "alternates": {"html": "https://wpnews.pro/news/solving-hack-the-box-challenges-with-gpt-5-6-luna-pro", "markdown": "https://wpnews.pro/news/solving-hack-the-box-challenges-with-gpt-5-6-luna-pro.md", "text": "https://wpnews.pro/news/solving-hack-the-box-challenges-with-gpt-5-6-luna-pro.txt", "jsonld": "https://wpnews.pro/news/solving-hack-the-box-challenges-with-gpt-5-6-luna-pro.jsonld"}}