rthng
- Karma
- 0
- Created
- ()
- Submissions
- 0
- Qwen3.8-27B at ~200 tok/s peak on an Apple M5 Max (twitter.com)
- Kimi 2.7 Code: 1T model with 1000 tokens per second (lithosai.com)
-
Either as a deployment platform or as an area that you are investigating for a tool/service. Or maybe you're just adding Azure integration to an existing product. I'd love to hear any opinions/experiences you might have…