GPT-5 Nano experienced an interruption over the last 24 hours, dropping its measured availability to 92.3% due to a bad_request error. The model subsequently recovered and returned to service, but it was the only evaluated endpoint to record an uptime failure during this monitoring period. Every other tracked model across OpenAI and Google maintained a 100% success rate across hourly checks.
Outside of the availability hiccup for GPT-5 Nano, overall performance held steady across latency and evaluation metrics. Gemini 2.5 Flash Lite recorded the fastest median response time overall, while Gemini 3.6 Flash and GPT-5 Mini shared the top position on the 7-day standing benchmark.
Speed
Response times across the 13 evaluated models spanned from sub-second latencies to over four seconds per request. Gemini 2.5 Flash Lite posted the lowest median response time at 648ms, followed closely by Gemini 2.5 Flash at 823ms and GPT-5.4 Mini at 993ms.
The table below lists the median response times and uptime recorded over the last 24 hours:
| Model | Median Response Time | Uptime |
|---|---|---|
| Gemini 2.5 Flash Lite | 648ms | 100% |
| Gemini 2.5 Flash | 823ms | 100% |
| GPT-5.4 Mini | 993ms | 100% |
| Gemini 3.1 Flash Lite | 1.05s | 100% |
| GPT-5.6 Terra | 1.17s | 100% |
| GPT-5.6 Luna | 1.18s | 100% |
| GPT-5.6 Sol | 1.30s | 100% |
| Gemini 3.5 Flash | 2.24s | 100% |
| Gemini 3.6 Flash | 2.54s | 100% |
| GPT-5 Nano | 2.78s | 92.3% |
| GPT-5 Mini | 3.07s | 100% |
| Gemini 3.1 Pro (preview) | 3.61s | 100% |
| Gemini 2.5 Pro | 4.32s | 100% |
GPT-5 Nano's median latency settled at 2.78s alongside its reduced uptime. At the slower end of the spectrum, Gemini 3.1 Pro (preview) registered a 3.61s median response time, while Gemini 2.5 Pro trailed the group at 4.32s.
Prices
Google's pricing structure for the Gemini family reflects distinct tiering across Lite, Flash, and Pro variants. Input costs range from $0.100 per million tokens up to $2.00 per million tokens, with output rates scaling up to $12.00 per million tokens.
Gemini 2.5 Flash Lite remains the most economical model in the monitored group at $0.100/M for input tokens and $0.400/M for output tokens. Gemini 3.1 Flash Lite is priced at $0.250/M input and $1.50/M output, slightly below Gemini 2.5 Flash at $0.300/M input and $2.50/M output.
Higher-tier models command larger premiums. Gemini 2.5 Pro costs $1.25/M input and $10.00/M output, while Gemini 3.5 Flash requires $1.50/M input and $9.00/M output. Gemini 3.1 Pro (preview) sits at the top of the pricing matrix, billed at $2.00/M for input tokens and $12.00/M for output tokens.
Benchmark
The 7-day standing benchmark saw flawless completion rates from Gemini 3.6 Flash and GPT-5 Mini, both scoring 100% of available points at $0 total spend. GPT-5.6 Terra earned third place with a 90.9% score, also at $0 spend.
Gemini 2.5 Flash and GPT-5.6 Luna tied at 84.9% of available points. However, running Gemini 2.5 Flash incurred $0.00912 in total spend, whereas GPT-5.6 Luna recorded $0 spend. Gemini 3.1 Flash Lite completed the monitored benchmark suite at 82.9% of points on a total spend of $0.00554.
These measurements reflect direct synthetic benchmark runs and automated hourly checks over the specified time windows. They prove specific API availability, latency, and task accuracy under test conditions, but they do not prove overall real-world user performance, long-term provider reliability, or private enterprise rate-limit behavior.
