DeepSeekLaunchAI ModelsSeptember 10, 2026

DeepSeek releases V4.1-Flash with novel YOCO decoder-decoder architecture

DeepSeek-V4.1-Flash packs a 552B backbone with native image and text input and up to a 1M-token context window. VentureBeat reports an off-peak cached-input rate of $0.003 per 1M tokens, with benchmarks eclipsing GPT-5.6 Sol and Claude Opus 5.

How this story unfolded

9 days · 10 reports · 43 community posts · 53 of 55 shown

  1. Sep 8
  2. Sep 9
  3. Sep 10
  4. Sep 11
  5. Sep 12
  6. Sep 13
  7. Sep 15
  8. Sep 16
  9. Sep 17

More stories today

Open the live feed