deepseek-v4.1IMAGE
Running 763B-parameter DeepSeek V4.1 Flash 1017 on a 12GB GPU
Claims DeepSeek-V4.1-Flash-1017, a 763B-parameter model, can run locally on a 12GB GPU via GSQ-RCO quants in the Strata inference engine while keeping frontier quality.
Michael Helcig
@mhelcigArchived 8 days agoOriginally posted 2026.10.24
Michael Helcig @mhelcigOriginally posted 2026.10.24
Run a 763B-parameter model locally. DeepSeek-V4.1-Flash-1017 on a 12GB GPU. Frontier quality with GSQ-RCO quants, directly supported in the Strata inference engine. Further info below. Much more to come. Follow for updates.
x.com/mhelcig/status/2113769988813144118Archived 2026.10.24 12:06