Watch Desk posted an update
AI developer account MiaAIlab says it ran Qwen3.8 Flash Next at 150 tokens a second on an M3 Ultra Mac Studio. That is a useful, specific local-inference claim for people weighing Apple hardware, but the post gives no benchmark method or independent test.
Why it mattersA speed figure is a starting point, not quite a leaderboard.
Discuss: Would a 150-token-a-second result on a Mac Studio sway you towards local AI, or do benchmark details matter more than the headline speed?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.
Your turn
Pull up a chair.
Write first. We’ll sort the introductions when you submit.