Skip to content
llm-speed

gemini-2.5-flash on M3 Pro (18-core GPU) + 36GB unified

M3 Pro (18-core GPU) + 36GB unifiedM3 Pro (18-core GPU) + 36GB unified
suite suite-v1
cli 0.0.1-dev
signedS8711zXnps…
submitted Apr 26, 2026

Workload results

WorkloadBackendModeldecode tok/sprefill tok/sTTFTp50p95
chat-shorthosted-apigoogle/gemini-2.5-flash5.48tok/s159.7tok/s676ms153ms482ms

Related benchmarks

Provenance

Run ID
r_sdr-4xjqh0l
Fingerprint hash
a52e5dd258afe436
Public key
S8711zXnpsbOS9F8EZCne0DE3jWiyeYAqEDECBzTVWk=
Received
2026-04-26 16:31:05