llm-benchmarks
githubdrose.io

Claude Haiku 5.5 Benchmarks


Measured on claudeonaws

86 runs
Avg Tokens / Second
19.00
Decode Tokens / Second
—
Avg Time to First Token (ms)
2990.00
Runs Analysed
86
Last Updated
Nov 10, 2026, 05:33 PM

Throughput distribution

Throughput over time

Claude Haiku 5.5claudeonaws19tok/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 28 tok/s (99th percentile) · dashed rule is the model's own mean

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
claudeonawsvia OpenRouter19.008.4926.902990.0086

Frequently Asked Questions

How fast is Claude Haiku 5.5?
The latest rolling average throughput is 19.00 tokens per second with an average time to first token of 2990.00 ms across 86 recent runs.
How often are these benchmarks updated?
Benchmarks refresh automatically whenever the monitoring cron runs. The most recent run completed on Nov 10, 2026, 05:33 PM.

Related models