Benchmark Overview
Avg Tokens / Second
13.90
Avg Time to First Token (ms)
1700.00
Runs Analysed
173
Last Updated
Mar 21, 2026, 03:02 AM
Key Insights
gpt-5.2-codex streams at 13.90 tokens/second on average across the last 173 benchmark runs.
Performance fluctuated by 24.76 tokens/second (178.1% coefficient of variation), indicating variable behavior across benchmark runs.
Average time to first token is 1700.00 ms (moderate latency), consider alternatives for latency-sensitive workloads.
Latest measurements completed on Mar 21, 2026, 03:02 AM based on 173 total samples.
Benchmark Samples
| Provider | Model | Avg Toks/Sec | Min | Max | Avg TTF (ms) |
|---|---|---|---|---|---|
| openai | 13.90 | 1.94 | 26.70 | 1700.00 |
Frequently Asked Questions
The latest rolling average throughput is 13.90 tokens per second with an average time to first token of 1700.00 ms across 173 recent runs.
Benchmarks refresh automatically whenever the monitoring cron runs. The most recent run completed on Mar 21, 2026, 03:02 AM.