LaunchAI ModelsAugust 3, 2026

DeepSeek V4 Flash launches with vision support

DeepSeek-V4-Flash-0731, a 284B-parameter MoE model, is now live with vision support. It scores 82.7 on Terminal-Bench 2.1, ahead of V4 Pro Preview at 72.1, and reportedly completes benchmark tasks at 105x lower cost than Fable 5. Priced near 2 cents per million input tokens.

How this story unfolded

4 weeks · 6 reports · 23 community posts · 29 of 31 shown

Aug 1
Aug 2
Aug 3
Aug 4
Aug 5
- 🤯 DeepSeek reports V4 Flash at 82.7 on Terminal-Bench 2.1, ahead of V4 Pro Preview at 72.1, despite...
Aug 20
- The boring way to run Deepseek V4 Flash-0731 130-150 tks - 16x5060ti 16GB over 2 PLX88096 switches
Aug 24
- deepseek-v4-flash-0731 - surprisingly usable
Aug 25
- Almost 20k downloads so far on @huggingface The GGUF quantizations of...
Aug 28
- Quick reminder: in 4 hours you will be able to download and run sota AI. If you have enough compute...
Aug 30
- 67-84 t/s DeepSeek flash v4 off 2x GX10s
Aug 31

AI Models by email

Get an email when there's news on AI Models

No news that day, no email.

DeepSeek V4 Flash launches with vision support

How this story unfolded

AI Models by email

More stories today