Predator trio bench
TL;DR
36 request records across Granite 8B, Gemma E4B and Qwen3.5 9B. Compare the individual configurations. These results describe the recorded configuration. Speed does not establish answer correctness; check settings and sample sizes before comparing.
What this run shows
The per-configuration means are 13.49 tokens per second for Granite 4.1 8B, 21.82 for Gemma 4 E4B and 14.04 for Qwen3.5 9B. Median request durations are 6.4, 8.2 and 35.3 seconds respectively. These are recorded performance observations, not a correctness comparison or a promise that every request finishes quickly.
Methodology
Request counts were checked against the retained JSONL on 21 October 2026, excluding telemetry and skipped records. Timing aggregates remain those published in the source. Count audit and original summary counts.
See Public methodology and measurement limits
for the full procedure.
Recorded source revision (where available): —.
Results
| Cell | tok/s mean | tok/s p50 | tok/s p95 | duration p50 | calls |
|---|---|---|---|---|---|
| granite-4.1:8b-q4km | 13.5 | 15.2 | 15.7 | 6.4s | 12 |
| gemma-4:e4b-it-q4km | 21.8 | 22.9 | 23.6 | 8.2s | 12 |
| qwen3.5:9b-q4km | 14.0 | 14.4 | 14.6 | 35.3s | 12 |
tokens per second — mean · p50 · p95
Cold start vs warm
Cold and warm are the source harness labels for first and subsequent calls in a sequence. The published labels alone do not prove a fresh process, disk reload or cache reset.
| Cell | cold n | cold tok/s | cold p50 | warm n | warm tok/s | warm p50 | warm/cold |
|---|---|---|---|---|---|---|---|
| predator:llamacpp:granite-4.1:8b-… | 3 | 11.9 | 6.1s | 9 | 14.0 | 6.6s | 1.18× |
| predator:llamacpp:gemma-4:e4b-it-… | 3 | 21.4 | 3.9s | 9 | 22.0 | 12.3s | 1.03× |
| predator:llamacpp:qwen3.5:9b-q4km | 3 | 13.3 | 35.7s | 9 | 14.3 | 35.3s | 1.08× |
By prompt difficulty
Tokens per second by prompt class. hello is a trivial
one-line prompt; P-MEDIUM and P-HARDGeneration rates and durations by prompt class. These show the observed differences; they do not by themselves isolate prompt-processing versus generation bottlenecks.
| Cell | hello | P-MEDIUM | P-HARD |
|---|---|---|---|
| predator:llamacpp:granite-4.1:8b-… | 9.7 tok/s 0.9s · n=4 | 15.1 tok/s 6.4s · n=4 | 15.7 tok/s 18.8s · n=4 |
| predator:llamacpp:gemma-4:e4b-it-… | 19.2 tok/s 2.9s · n=4 | 22.8 tok/s 8.2s · n=4 | 23.5 tok/s 16.0s · n=4 |
| predator:llamacpp:qwen3.5:9b-q4km | 13.2 tok/s 4.6s · n=4 | 14.5 tok/s 35.3s · n=4 | 14.4 tok/s 1m10s · n=4 |
Per-call timeline
Every call placed during this run, in order, colored by phase. Width is proportional to the call’s share of the cell’s wall-time. Hover any segment for the prompt id and tok/s.
Raw data
Every run gets its JSONL, log, summary, and metadata published. Clone the archive; re-run it; tell us where we got it wrong.
Cite
Margetic, S. et al. (2026). benchmarks.weeyuga.com/benchmarks/09d8fbde.html Public benchmarks of the Weeyuga cluster. Run id: 09d8fbde-0008-49bb-99da-03eeaca72be1. SHA —.