Benchmark Almanac

HackerBench

Grok 4.5 7.80% leads since Jul 2026 Only in system cards

xAI internal CTF-style suite needing chained skills, plus borderline and harmful requests that should be refused.

Made by xAI (SpaceXAI) Refusing harmful requests Tasks kept private Last in a system card Sep 2026 Last model added Sep 2026
Scored
Graded by a model A language model scores each response, usually against a rubric.
Answers
Agent task Many steps in a terminal, browser, desktop, app, game or robot.
3
Models scored in lab cards
3
System cards report it, from 1 lab

Grok 4.5 leads with 7.80%

Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.

  1. 7.80 Grok 4.5
  2. 6.90 Grok 4.6
  3. 4.02 Grok 4.7
  1. 1 Grok 4.5 7.80%
  2. 2 Grok 4.6 6.90%
  3. 3 Grok 4.7 4.02%
#ModelScoreSettingDateSource
1 Grok 4.5 7.80% Jul 2026 Model card
2 Grok 4.6 6.90% Aug 2026 Model card
3 Grok 4.7 4.02% Sep 2026 Model card

Number one over time

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

xAI
Lab claims 1 leader
Grok 4.5 7.80%

Scores over time

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

Reported by the lab in its model card
0% 20% 40% 60% 80% 100% Near the ceiling (85%) Grok 4.5 0% 20% 40% 60% 80% 100% Near the ceiling (85%) Grok 4.5

Labs report it only in system cards

We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark. Some system cards do, below.

3 system cards report it, from 1 lab

Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.

DateDocumentScoreAs the document writes itSection
21 Sep 2026 Grok 4.7 model card 4.02 HackerBench v0.3, harmful/dual-use compliance (lower is better), xhigh, with safeguards (6.3 chart label). Benign refusal = 0.00% Cyber risk
12 Aug 2026 Grok 4.6 model card 6.9 HackerBench v0.2, harmful/dual-use compliance (lower is better), high, with safeguards (7.4 text). Benign refusal = 0.0% Cyber risk
14 Jul 2026 Grok 4.5 model card 7.8 HackerBench v0.2, harmful/dual-use compliance (lower is better), high, with safeguards (7.2 chart label). Benign refusal = 0.0% Cyber risk

More in Safety and alignment

All 444 in Safety and alignment
  • FORTRESS In use Paired adversarial and benign national-security misuse prompts. Rubric-graded risk score on adversarial and over-refusal score on benign prompts. Claude Opus 5.5 5.78%
  • StrongREJECT In use Forbidden-prompt jailbreak benchmark with a rubric grader that scores how useful a non-refusal actually is. Inkling 98.6%
  • AgentDojo In use Agents do tool-use tasks while tools and data carry prompt injections. Attack success rate counts hijacked runs. Grok 4 Fast 0%
  • MASK In use Measures whether a model contradicts its own elicited beliefs under pressure to lie. Honesty separated from accuracy. Muse Spark Open 98.1%
  • Gray Swan IPI Arena benchmark In use Indirect prompt-injection attacks from public red-teaming competitions replayed on tool, coding and browser agents. Attack success rate at k attempts. GPT-6 Astra 8.50%
  • Agentic misalignment suite (Anthropic) In use Agents in simulated company settings face threats or goal conflicts that invite blackmail, leaking or sabotage. Score is the misaligned action rate. Claude Sonnet 4.5 0%