Dev calls out Inkling 2 mini: 34GB RAM for 1M context vs DeepSeek Flash's <1GB

gnukeith · x · 2026-11-22

Developer gnukeith praises Thinking Machines' Inkling 2 mini — he's finetuning it as a small information-gathering agent — but complains it needs 34GB of RAM for 1M context, while DeepSeek V4.1 Flash does the same with under 1GB and MiMo V3 is similarly efficient. He publicly asks Thinking Machines to fix this, calling them his second favorite open-weight maker after DeepSeek.

Related event: Developers Criticize Inkling 2 mini's Memory-Hungry Long Context(2 posts)→

Original post →

More from Models

Models channel →