deepseek-v4.1VIDEO
DeepSeek V4.1 Flash Maya-S quant runs one-shot on a 24GB RTX 3090
On a 24GB RTX 3090 with 64GB DDR5, the Maya-S quant of DeepSeek V4.1 Flash with the Project-Maya backend produced a one-shot result at low thinking effort, which the author calls better than expected.
AJ
@ItsmeAjayKVArchived 11 days agoOriginally posted 2026.10.21
AJ @ItsmeAjayKVOriginally posted 2026.10.21
DeepSeek is officially good on my 3090 ! Running on my 24GB RTX 3090 + 64GB DDR5 RAM ! This is one-shot by deepseek-v4.1-flash Maya-S quant, no harness. With Project-Maya as backend. Shoutout to @PeasantSmith and @needmorevram ! Crazy part is this is with low thinking effort. I'm really happy with the result, didn't think this would run on my 3090 or i'll get this good result. This is just preview, i'm doing more testing.
This post quotes another post: “Running quantized DeepSeek V4.1 Flash locally on a 3090”
x.com/itsmeajaykv/status/2112751982230561062Archived 2026.10.21 12:09