Leak: Gemini 3.5 Pro with 2M-token context and Deep Think mode

Skip to content

Alex Getman @alexgetmancom · 5 Jul 2026, 00:07 UTC

LEAK: Gemini 3.5 Pro will get a 2M-token context window

Plus Deep Think mode for complex code, logic, architecture, and legal tasks.

Pricing is still unofficial:

$12-15 per 1M input tokens
$36-45 per 1M output tokens

Pro is expected to be 8-10x more expensive than Flash.

Infographic on Gemini 3.5 Pro: 2 million tokens, Deep Think mode for code, logic, architecture, legal tasks, unofficial pricing $12-15 per 1M input and $36-45 per 1M output tokens, 8-10x more expensive than Flash.

Русская версия Markdown

Analysis

AI API pricing: calculate the cost of a finished task

Related posts

12 NovGemini 4 Hydrogen beats Argon in the Artificial Analysis intelligence index at the standard High reasoning level (51 vs 50), and at the max Xhigh level it hits 54.

12 NovGemini 4 benchmarks are here Pretty solid model for everyday work, as long as it doesn't hallucinate as much as Argon.

12 NovGemini 4 Hydrogen is out!

12 NovGemini 4 Hydrogen is launching today. It's a cheaper model than Carbon, so logically, it should compete with GPT Sol.