COMMUNITY BENCHMARKS LIVE
Real numbers from
real Macs.
Measured by the people who own them. Every submission is one locked run — two prompt sizes, five rounds after a discarded warmup, greedy, prefix cache off, one request at a time — which is the only reason two people's numbers can be compared.
One board: rapid-mlx benchmark run + share
puts your result on it, next to everyone else with the same Mac.
Runs from the old rapid-mlx bench --submit benchmark
fill cells nobody has measured yet, badged with their version and
never mixed into current medians.
What runs well on which Mac
Loading community measurements…
Latest community runs
Recent LLM, image and video runs from
rapid-mlx benchmark run + share, newest
first. The board above is built from these same runs.
Representative community results
Latest contributions
Community atlas
Every character is a real Mac adding another point to the map.
Share a run from your Mac
Run locally first. Sharing is a separate command that previews the exact payload and asks for consent before anything leaves your Mac.
Where it lands: the board
above, in your Mac's row, and on your contributor page —
share prints the link to your result.
Published: chip, RAM, core counts, OS and package versions, the model alias, and the timings.
Never sent: prompts, files, or any hardware serial. The
id behind your board name is random per install and
resets when you delete ~/.rapid-mlx/bench-install-id.
Never stored: your IP — it is used transiently for abuse limits.