deepseek-v4.1POST
DeepSeek V4.1-Flash-1017 vs Opus 5: wins DeepSWE, loses Terminal-Bench
On Hugging Face under MIT, it beats Opus 5 on DeepSWE 77 vs 74 but trails on Terminal-Bench 4.0 at 42.3 vs 51.8 — a big jump, not yet Opus level.
doanbactam
@m0nle0zArchived 10 days agoOriginally posted 2026.10.17
doanbactam @m0nle0zOriginally posted 2026.10.17
deepseek put V4.1-Flash-1017 on hugging face, MIT license. beats Opus 5 on DeepSWE (77 vs 74), loses on Terminal-Bench 4.0 (42.3 vs 51.8). big jump, not Opus level yet
x.com/m0nle0z/status/2111375640817750217Archived 2026.10.18 16:45