Open-source Strata runs DeepSeek V4.1 on an RTX 3060 at 9-11 tokens/sec
alex_verem · x · 2026-11-08
Strata is a free, open-source desktop app (MIT-licensed, Windows/Mac/Linux, no account needed) that runs large AI models locally — including DeepSeek V4.1 on a 2021 gaming GPU.
- Key trick: model weights that don't fit in memory are streamed straight off the SSD.
- On an RTX 3060 it generates 9-11 tokens/sec, faster than reading speed; requires 46GB RAM and 265GB free disk.
- Less powerful machines can run smaller models like Qwen3.8 through the same app.
- Supports chat, coding and image work, and integrates with favorite coding tools.
Related event: Open-source Strata runs DeepSeek locally on an RTX 3060 at 9-11 tokens/s(2 posts)→
More from Infra
- ResearchBoost Agent speeds up real research workloads 2.51x on average across 12 tasks — YouJiacheng · 2026-11-08
- Dev builds Cray-2 emulator from scratch, runs a GPT-3-class LLM at 1 token per 4s — inductionheads · 2026-11-08
- Hyperbolic CEO on compute market-making: bridging 3-year contracts and 3-week needs — YiMaTweets · 2026-11-08
- Ben Bajarin mocks 'we don't need data centers' take: CSPs were 'useless' too — BenBajarin · 2026-11-08
- KV Cache Quantization Tested Across 35 Quants: 8-bit Is Essentially Free for Qwen3.8 27B — Hefty_Wolverine_553 · 2026-11-08
- Developer burns $4,600 on rented GPUs with zero training steps, finishes same RL run on Tinker for under $100 — simonguozirui · 2026-11-08