llm-benchmarks
githubdrose.io

xai Provider Benchmarks


Measured over the last 30 days

Models Tracked
14
Avg Tokens / Second
9.90
Avg Time to First Token (ms)
13485.33
Last Updated
Sep 1, 2026

All xai models

14 tracked · fastest first
ModelAvg Toks/SecMinMaxAvg TTF (ms)
Grok 4.2034.7025.9042.50820.00
Grok 4.2034.6019.2052.30970.00
Grok 4.310.606.4713.805530.00
Grok 4.310.506.2615.505670.00
Grok 4.57.585.1013.907070.00
Grok 4.56.203.608.878700.00
Grok Build 0.14.301.847.5915600.00
Grok 4.20 Multi-Agent3.971.1932.5022800.00
Grok Build 0.13.592.305.8318000.00
Grok 4.20 Multi-Agent3.462.275.9018900.00
Grok 4.63.251.219.3922500.00
Pareto Code Router2.732.732.7321500.00
Grok Latest2.701.076.1725200.00
Grok 4.62.671.324.1125600.00

Frequently Asked Questions

Which xai model is fastest?
Based on recent tests, Grok 4.20 shows the highest average throughput among tracked xai models.
How many recent measurements feed this dashboard?
This provider summary aggregates 381 individual prompts measured across 334 monitoring runs over the past month.