Benchmark Almanac

Claude Mythos 5

Claude Mythos 5 ranks first on 2 of the 4 benchmarks with five or more models scored.

Anthropic Released June 9th, 2026 Family Claude Mythos The lab's model card Compare with others
4
Benchmarks with a score
2
Ranked first, out of five or more models
2
In the top three
Anthropic Jun 2026

Claude Mythos 5

Closed weights Flagship

No Epoch Capabilities Index score yet
  • ExploitGym 28.4% #6 /13
  • dfbench 69% #1 /18
  • CryptanalysisBench 34.2% #1 /5

Where has it been number one?

Claude Mythos 5 led 3 benchmarks at some point, for a median of 112 days. It last lost ExploitGym to GPT-5.6 Sol on July 9th, 2026. It still tops 2 that nobody has tested since.

Source: Official leaderboards

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Safety and security

BenchmarkScoreRankSource
dfbench 69% #1 of 18 Official leaderboard
CryptanalysisBench 34.2% #1 of 5 Official leaderboard
ExploitGym 28.4% #6 of 13 Official leaderboard

Other

BenchmarkScoreRankSource
Cyber ECI 168.3 #5 of 65 Official leaderboard

What does it cost to run?

On the one leaderboard that publishes a price, its best run costs 12 times the median model's. No model comes close to its score for less.

Benchmark Score Price From cheapest to priciest Against the median Close to its score for less Close for less
dfbench 69% $99.19 12× the median 12× None

Who measured it?

Independent runs cover all 4 of its benchmarks.

  • An independent run only 4
Official leaderboards scored it on 4.

Sources: Official leaderboards, Epoch AI (CC BY 4.0) and labs’ model cards