Compare
Compare local AI hardware with workload-aware output.
AMD Instinct MI325X 256GB wins for coding in balanced mode
Based on model fit, speed, and quality across top recommendations.
AMD Instinct MI325X 256GB
WinnerSDeepSeek V4 Flash
llama.cppq4-k-mRuns well
201.1 GB / 256.0 GB
82.5 tok/s686K ctx
SDevstral 2 123B Instruct
llama.cppq8-0Runs well
163.5 GB / 256.0 GB
39.8 tok/s256K ctx
SMiniMax M2.7
llama.cppq4-k-mRuns well
170.6 GB / 256.0 GB
101.4 tok/s205K ctx
Quick comparison
| Metric | AMD Instinct MI325X 256GB | RTX 4090 24GB |
|---|---|---|
| Models that fit | 3 | 3 |
| Avg decode tok/s | 74.6 | 37.9 |
| Best grade score | 99 | 93 |
