Ox Alpha is scored %63 on the DeepSWE bench run by @davis7. for | Tech Twitter

Finding signal on Twitter is more difficult than it used to be. We curate the best tweets on topics like AI, startups, and product development every weekday so you can focus on what matters.

Trending//Time Travel//Threads//Articles//Topics//Profiles//Newsletter//Developers

Focus

Press Space to continue

Built by Daniel Hunter

Ox Alpha is scored %63 on the DeepSWE bench run by @davis7. for | Tech Twitter

Press Space to continue

Ox Alpha is scored %63 on the DeepSWE bench run by @davis7. for reference: > DeepSeek V4 Pro: 63% > Grok 4.6: 65% > Gemini 3.7 Flash: 65% if this is actually a GLM flash or a small open weight model that can be run on 1-2 DGX Sparks, that’s crazy value for compute. https://x.com/Hesamation/status/20913052… https://twitter.com/davis7/status/209128…

Content

Previous tweet Next tweet More from @hesamation