Model Releases
I don't really ever trust benchmarks, so ocassionally I'm stress-vibe-testing a bunch of new models on some very complex agent work (hundred…
I don't really ever trust benchmarks, so ocassionally I'm stress-vibe-testing a bunch of new models on some very complex agent work (hundreds of tools, not your simple coding agent stuff) and to my su
I don't really ever trust benchmarks, so ocassionally I'm stress-vibe-testing a bunch of new models on some very complex agent work (hundreds of tools, not your simple coding agent stuff) and to my surprise, the clear winner is DeepSeek V4 Pro ...which is also the cheapest!
Source: Clem Delangue (X) | 2026-05-07