Compare
Compare local AI hardware with workload-aware output.
RTX 3090 24GB wins for coding in balanced mode
Based on model fit, speed, and quality across top recommendations.
RTX 3090 24GB
WinnerSCodestral 2 25.08
llama.cppq4-k-mRuns well
19.2 GB / 24.0 GB
46.9 tok/s48K ctx
SQwen 3.6 27B
llama.cppq4-k-mTight fit
20.7 GB / 24.0 GB
28.8 tok/s69K ctx
SDevstral Small 2 24B Instruct
llama.cppq4-k-mTight fit
20.4 GB / 24.0 GB
48.1 tok/s40K ctx
Quick comparison
| Metric | RTX 3090 24GB | RTX 4090 24GB |
|---|---|---|
| Models that fit | 3 | 3 |
| Avg decode tok/s | 41.3 | 37.9 |
| Best grade score | 93 | 93 |