Those 27.3% are still in the ballpark of modern models:

- Sonnet 5 - 12.4%

- Luna - 17.3%

- Grok 4.6 - 20.3%

- Sol - 37.3%

- GLM 5.3 - 41.8%

- Opus 5 - 51.8%

p1esk 39 minutes ago | parent | next [-]

Astra is 58%. The current title says it's "rivaling Astra"

	▲	thereitgoes456 33 minutes ago \| parent [-]
		It is rivaling Astra, on their own benchmark that they made (FrontierCode), that they ran themselves in their own closed-source ecosystem that isn’t reproducible by anyone.

Readerium 2 hours ago | parent | prev [-]

DeepSeek v4.1 Flash 31.2%

	▲	nicce 24 minutes ago \| parent [-]
		I have been using it today the whole day and it is definitely better than Luna. Great that bench agrees.