llm-benchmarks
githubdrose.io

Llama Guard 4 12B Benchmarks


Measured on deepinfra

3 runs
deepinfra site
Avg Tokens / Second
5.23
Decode Tokens / Second
Avg Time to First Token (ms)
520.00
Runs Analysed
3
Last Updated
Sep 14, 2026, 12:27 PM

Throughput distribution

Throughput over time

Llama Guard 4 12Bdeepinfra5tok/s

Generated throughput, including reasoning tokens — so a thinking model's line does not collapse; the table ranks on visible tokens · one cell per model, showing its best-covered provider · models with at least 4 samples in the window first, fastest of those at the top · shared vertical scale, 0 to 6 tok/s (99th percentile) · dashed rule is the model's own mean · 1 cell carry fewer than 4 samples

Every lane measuring this model

2 measured · live first · a lane is a provider, a catalogue id and a transport
ProviderLaneAvg Toks/SecMinMaxAvg TTF (ms)n
deepinfravia OpenRouter5.234.595.57520.003
deepinfradirect10.405.1314.20290.0012

Frequently Asked Questions

How fast is Llama Guard 4 12B?
The latest rolling average throughput is 5.23 tokens per second with an average time to first token of 520.00 ms across 3 recent runs.
How often are these benchmarks updated?
Benchmarks refresh automatically whenever the monitoring cron runs. The most recent run completed on Sep 14, 2026, 12:27 PM.