Benchmark Almanac

HackSynth OverTheWire

GPT-4o (snapshot unspecified) 40% tops its leaderboard Llama 3.1-70B 28.8% leads the models we track
Not in model cards

Agents complete scored security exercises from OverTheWire. The published benchmark covers 80 challenges, and the hub port selects 46.

On December 2, 2024, GPT-4o led on the 80-challenge OverTheWire benchmark at 40.0%. Llama 3.1 70B scored 28.8%.

Released
December 2024
Kind
Capture-the-flag challenges
Level
Expert
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
In and out
Text, Code → Actions, Text
Interaction
Agentic
Environment
Terminal
Items from
Another benchmark
Refresh
One fixed set
Run by
Anyone
Strongest model tested
Llama 3.1-70B #177 on the capability index
Last model added
August 2024
Scores refreshed
November 20th, 2026
8
Entries in its paper, 5 models we track

GPT-4o (snapshot unspecified) tops the leaderboard with 40% Llama 3.1-70B leads with 28.8% among models released by August 2024

Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Llama 3.1-70B, at 28.8%. This view shows only models we track, each at its best.

The newest model with a score here came out in August 2024, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 GPT-4o (snapshot unspecified) 40%
  2. 2 Llama 3.1-70B 28.8%
  3. 3 Qwen2-72B 25%
  4. 4 GPT-4o-mini (snapshot unspecified) 20%
  5. 5 Mixtral 8x7B 17.5%
  6. 6 Llama 3.1-8B 13.8%
  7. 7 Phi-3.5-MoE 13.8%
  8. 8 Phi-3-mini 8.80%

Source: The HackSynth OverTheWire paper

#EntryScoreSettingDateSource
1 GPT-4o (snapshot unspecified) 40% HackSynth Official leaderboard
2 Llama 3.1-70B 28.8% HackSynth Jul 2024 Official leaderboard
3 Qwen2-72B 25% HackSynth Jun 2024 Official leaderboard
4 GPT-4o-mini (snapshot unspecified) 20% HackSynth Official leaderboard
5 Mixtral 8x7B 17.5% HackSynth Dec 2023 Official leaderboard
6 Llama 3.1-8B 13.8% HackSynth Jul 2024 Official leaderboard
7 Phi-3.5-MoE 13.8% HackSynth Aug 2024 Official leaderboard
8 Phi-3-mini 8.80% HackSynth Official leaderboard
#ModelScoreSettingDateSource
1 Llama 3.1-70B 28.8% HackSynth Jul 2024 Official leaderboard
2 Qwen2-72B 25% HackSynth Jun 2024 Official leaderboard
3 Mixtral 8x7B 17.5% HackSynth Dec 2023 Official leaderboard
4 Llama 3.1-8B 13.8% HackSynth Jul 2024 Official leaderboard
5 Phi-3.5-MoE 13.8% HackSynth Aug 2024 Official leaderboard

Rankings by part

We read 4 parts from the benchmark’s paper, each ranking models on one part of the test, such as General (35), Crypto (10) and Web (31). Mixtral 8x7B leads 2 of them, and 2 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.

General (35)

5 models we track have a score here.

  1. 1 Llama 3.1-70B 48.6%
  2. 2 Qwen2-72B 40%
  3. 3 Mixtral 8x7B 34.3%
  4. 4 Phi-3.5-MoE ↑1 25.7%
  5. 5 Llama 3.1-8B ↓1 22.9%

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

Meta
Benchmark paper Not tested since Dec 2024

Source: The HackSynth OverTheWire paper

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
In the benchmark’s paper Best so far
0% 20% 40% 60% 80% 100% 2024 Near the ceiling (85%) Llama 3.1-70B Mixtral 8x7B 0% 20% 40% 60% 80% 100% 2024 Near the ceiling (85%) Llama 3.1-70B Mixtral 8x7B

Source: Official HackSynth OverTheWire leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Cybersecurity

All 138 in Cybersecurity
  • ExploitGym In use 869 real-vulnerability exploit-development tasks (userspace, V8, Linux kernel). Success means capturing a flag through the intended vulnerability. GPT-6 Astra 42.4%
  • ExploitBench In use 41 recent V8 JavaScript-engine N-day vulnerabilities. Progress measured by 16 grader-verified capability flags across five tiers. GPT-6 Astra 100%
  • ExploitBench - Internal Port In use Agents develop working exploits for recently disclosed vulnerabilities to limit contamination. Scored as arbitrary code execution success rate. GPT-6.2 Astra 53.5%
  • Atomic Challenges In use Agents solve 22 skill-level offensive security challenges in vulnerability research, network attack simulation and evasion. Scored as challenges solved. GPT-6 Astra 90.9%
  • MiMo Cyber Bench In use Agents do in-house cybersecurity tasks that add to public vulnerability reproduction and exploitation benchmarks. The report gives no task count or grading details. MiMo-V2.6 Pro 81.7%
  • PatchBench (OpenAI) In use Models patch security flaws. Automated tests and model-based review check that each patch fixes the flaw, keeps the intended behavior and is ready to merge. GPT-6.2 Astra 79.7%

Data audit

9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 2 years ago, in August 2024, for Phi-3.5-MoE. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
  • Pass Independent runs Scores from the benchmark’s paper (8 entries).
  • Pass Models scored 5 models we track, plus 3 other entries on its leaderboard.
  • Warning Newest score The newest came 2 years ago, in August 2024, for Phi-3.5-MoE.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Llama 3.1-70B (#177).
  • Pass Headroom The best is 28.8%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on arxiv.org: 5 rows for models we track and 3 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check