Probed: partially recovered the spoken codeword. Treat sound commentary as unverified.
Gemini 4 Argon is Google's flagship reasoning model, built for complex software engineering, deep mathematical and scientific problem solving, and autonomous multi-step agents. Google trained it to keep context across...
| Field | Per token (USD) | Per million |
|---|---|---|
| prompt | 0.000001 | $1.00 |
| completion | 0.000005 | $5.00 |
| image | 0.000001 | $1.00 |
| audio | 0.000001 | $1.00 |
| input_audio_cache | 0.00000005 | $0.050 |
| web_search | 0.014 | $14000 |
| internal_reasoning | 0.000005 | $5.00 |
| input_cache_read | 0.00000005 | $0.050 |
| input_cache_write | 0.0000001875 | $0.188 |
{
"temperature": null,
"top_p": null,
"top_k": null,
"frequency_penalty": null,
"presence_penalty": null,
"repetition_penalty": null
}