All systems operationalGPT-6 Astra is live
Gemini 3.8 Flash at its default thinking level.
thinking_level=MEDIUM, the default for this model.
Latency
1.9s
Context
1.05M
Input
$0.75/M
Output
$3.75/M
{
"model": "gemini-3.8-flash-medium",
"messages": [
{
"role": "user",
"content": "Hello, world."
}
]
}