Qwen3.6 27B + 35B on vLLM, single R9700 (gfx1201)
Reddit user shares vLLM tuning results running Qwen3.6 27B and 35B on a single Radeon AI Pro R9700 (gfx1201), reporting they're "pretty happy" with the performance. Claude helped configure the setup, and the user is already looking forward to Qwen3.8.
1 source
Daily brief
Get tomorrow's AI brief in your inbox
More stories today
- New tutorial: build a web browsing agent with Stagehand v4
- OpenAI Build Week participants built projects with Codex
- Tool computes differential inverse kinematics using MuJoCo
- Run interactive IDEs on Amazon EKS with SageMaker AI
- NVIDIA releases open-weights Magpie TTS for multilingual voice agents