llm-benchmarks
githubdrose.io

Hermes 4 405B Benchmarks


Measured on nebius

1 runs
Avg Tokens / Second
17.00
Decode Tokens / Second
—
Avg Time to First Token (ms)
950.00
Runs Analysed
1
Last Updated
Nov 5, 2026, 10:53 AM

Throughput distribution

Throughput over time

Hermes 4 405Bnebius17tok/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 18 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
nebiusdirect17.0017.0017.00950.001
nebiusvia OpenRouter30.0028.2031.80660.002

Frequently Asked Questions

How fast is Hermes 4 405B?
The latest rolling average throughput is 17.00 tokens per second with an average time to first token of 950.00 ms across 1 recent runs.
How often are these benchmarks updated?
Benchmarks refresh automatically whenever the monitoring cron runs. The most recent run completed on Nov 5, 2026, 10:53 AM.

Related models