DeepSeek Details V4.1-Flash: 552B MoE with Asymmetric 8B/16B Active Params, Beats Its Flagship
deepseek_ai · x · 2026-09-10
DeepSeek shared the architecture behind V4.1-Flash, key to its claimed flagship-beating results:
- 552B-parameter MoE with a novel Causal Encoder–Decoder architecture, highly asymmetric activation: only 8B active parameters for input, 16B for output
- Combined with new pre-training methods and larger-scale RL post-training, benchmark results surpass flagship models including DeepSeek-V4-Pro
More from Models
- MissionReport benchmark: only 3 of 540 AI-generated government reports pass all criteria — Exp_Mark · 2026-10-27
- Local beats cloud: Qwen 3.8 FN handles private student data as well as frontier models — dr_lm · 2026-10-27
- Aidan Gomez: model's CoT said no tool calls right before making them — aidangomez · 2026-10-27
- Nat Lambert: Cohere's model is narrow RL training on internal infra ill-suited for open weights — natolambert · 2026-10-27
- OpenAI teases DOTS is coming to Europe 'soon' — mark_k · 2026-10-27
- Clifford Pickover challenges AI to decipher a message he encoded in Norse runes — pickover · 2026-10-27