Why Mythos doesn't matter (for us)

Tenzai benchmarked GLM-5.2 against GPT-5.5 and Opus-4.8 on real exploit tasks. GLM wins on cost at $1.13/lead. Frontier models win on recall — but only at 2–3× the price.

Benchmarking GPT-5.6 Sol Medium, Sol Ultra, and Grok 4.5 on Chrome exploit development: after 2.096 billion tokens, only Sol Ultra produced a complete exploit chain.

Benchmarking GLM, DeepSeek, Kimi, and MiniMax against GPT, Claude, Gemini, and Opus on 15 real Fider vulnerabilities: frontier models still lead, but open-weight models are closing the gap fast.

XBOW benchmarks Grok 4.5 against Mythos, GPT-5.5, and GLM: it solves about 93% of benchmark vulnerabilities at fixed iterations and delivers the best solve rates in the $1–$10 range.

XBOW compares low-cost models GLM and Muse Spark 1.1 with Mythos and GPT-5.5 on real vulnerable applications, arguing the 'best' model depends on budget and workload.

XBOW asks whether frontier models like Mythos and GPT-5.5 are enough for defense: finding bugs is not coverage, and enterprise security needs persistence, validation, safety, and governance.