DeepSeek V4.1 Flash hits 74.2 on DeepSWE, beating Opus 5 and Gemini 3.8 Flash; asymmetric MoE revealed
nrehiew_ · x · 2026-09-10
nrehiew surfaced two striking details about DeepSeek V4.1 Flash:
- Architecture returns to encoder-decoder: 552B total parameters with asymmetric activation—8B active on decode, 16B on prefill. Different activation counts per phase within one model is a first.
- It tops the DeepSWE benchmark at 74.2, edging out Astra, Opus 5, and Gemini 3.8 Flash (all at 74).
Related event: Leaked DeepSeek V4.1 Benchmarks Point to New 552B Architecture(15 posts)→
More from Models
- DeepSeek V4.1 Flash is actually 748B params, safetensors analysis shows — DistanceSolar1449 · 2026-09-10
- Kimi K3 lands on RunPod: 2.8T params, 1M context, $3/$15 per 1M tokens — Kimi_Moonshot · 2026-09-10
- Dev burns 300M tokens on GLM 5.3 in a week and still has quota left — saibharadwaj · 2026-09-10
- Follow-up: a 3T-parameter model may already exist, scaling issues remain the wildcard — teortaxesTex · 2026-09-10
- Speculation: DeepSeek V4.1 Pro could be a 3.1T-param MoE with 2.6TB disk footprint — teortaxesTex · 2026-09-10
- New Book Teaches Beginners to Build and Fine-Tune Their Own GPT-Style SLMs, With Colab Notebooks — Roger_M_Taylor · 2026-09-10