4 ms·
I think the 'Agentic coding SWE-Bench Verified' [1] was actually the one benchmark where Google didn't even claim to beat Sonnet 4.5 ;-) [1] https://deepmind.g
by arendtio 11mo ago
I think the 'Agentic coding SWE-Bench Verified' [1] was actually the one benchmark where Google didn't even claim to beat Sonnet 4.5 ;-)
[1] https://deepmind.google/models/gemini/pro/ https://deepmind.google/models/gemini/pro/