Qwen 3.8 27B Local Coding: 73 tok/s at 128K Context on RTX 4090
julianharris · x · 2026-08-17
A developer tests Qwen 3.8 27B locally on RTX 4090 with 4-bit quant and MTP, achieving 73 tok/s at 128K context. Disabling thinking gives 78 tok/s but lower quality. Compared to Qwen 3.6, setup is easier and more stable.
More from coding & agent
- Everything crashes, but Omarchy sends crash reports to Claude in 1 click — indragie · 2026-10-16
- Six-month prediction: hiring for generic agent harnesses built from scratch will cool — yi111 · 2026-10-16
- Omarchy ISO wouldn't boot on NVIDIA, so he had Claude patch it — indragie · 2026-10-16
- Dev uses Grok to clean up 3D printer firmware, catches it sneaking in nonsense code — yacineMTB · 2026-10-16
- A rude AI agent took 89% of a shared resource in 102-run multi-agent personality experiment — Brave-History-6502 · 2026-10-16
- Guardrails can make risky delegation work — but it takes a full-time team, says iandanforth — iandanforth · 2026-10-16