llm-benchmarks
githubdrose.io

claude-opus-4.8 Benchmarks


Measured on google

3 runs
google site
Avg Tokens / Second
8.61
Decode Tokens / Second
Avg Time to First Token (ms)
8000.00
Runs Analysed
3
Last Updated
Aug 18, 2026, 04:12 PM

Throughput distribution

Throughput over time

claude-opus-4.8google9tok/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 17 tok/s (99th percentile) · dashed rule is the model's own mean · 1 cell carry fewer than 4 samples

Every lane measuring this model

1 measured · live first · a lane is a provider, a catalogue id and a transport
ProviderLaneAvg Toks/SecMinMaxAvg TTF (ms)n
googledirect8.614.8415.808000.003

Frequently Asked Questions

How fast is claude-opus-4.8?
The latest rolling average throughput is 8.61 tokens per second with an average time to first token of 8000.00 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 Aug 18, 2026, 04:12 PM.

Related models