Same Mac. Same model. Same words. My 512 GB M3 Ultra ran GLM-5.3-Flash on MLX last week: 27 tok/s…
@volatilemarkts2.6K followers
28,183 views
11 Comments
Comments from the original post on X (Twitter)

@ashxhart
Let’s get that to 100tps 👀
4 likes

@thybead
61 tok/s on the same box is a nice jump. does it hold once the context gets ugly?
2 likes

@volatilemarkts
Like duct tape.
1 like

@1x3e2
Yep, it almost doubled the speed on my sparks with the same model. And I have already seen improvements with the new updates and recipes. Excited to see where it goes with so many people being able to find these new techniques.
2 likes

@pandresgq
The eight-request result is the real systems datapoint: concurrency exposes memory-bandwidth headroom that single-stream tok/s hides.
1 like

@MiaAI_lab
🔥🔥
1 like

@aienscore
2.25x on the same box, same model, measured with your own probe, is hard to argue with. The first token drop, 0.46 to 0.20, is the part that gets me. That's where interactive use lives or dies.

@MichaelGannotti
Love it.

@brandon_lutz
With vision?

@blockbrain_labs
Crushing

@AndriusStepona1
mlx 27 then tf 61 and eight concurrent still matches solo byte for byte
