Everything about Gemini 3 Pro Preview

Overview

Google DeepMind released Gemini 3 Pro Preview on November 18, 2025. It has no declared reasoning mode: it answers directly, without an extra billed thinking step. Its weights are not published, so it only runs remotely, through the Google DeepMind API or partner hosts (10 listed on models.dev). That makes it 11 months old as of October 22, 2026.

We file it under frontier models (24 models in this category on the site).

What do independent benchmarks say?

According to results published by Epoch AI (checked on October 22, 2026), Gemini 3 Pro Preview has a capability index of 87 out of 100, computed on 4 benchmarks from our basket.

That puts it at #14 in our weekly model ranking.

• GPQA Diamond (PhD-level science questions): 92.6% — the best tracked model, Gemini 4 Argon, reaches 96.4%
• OTIS Mock AIME (competition mathematics problems): 91.4% — the best tracked model, GPT-5.5 Pro, reaches 100%
• SWE-bench Verified (fixing real software bugs): 72.9% — the best tracked model, Claude Opus 4.7, reaches 83.5%
• Terminal-Bench (agent tasks in a terminal): 69.4% — the best tracked model, GPT-5.5, reaches 84.7%

How much does Gemini 3 Pro Preview cost?

Gemini 3 Pro Preview costs $2 per million input tokens and $12 per million output tokens (median of 10 providers, checked on October 22, 2026).

In practice, an agent task that reads 20,000 tokens and writes 2,000 costs about $0.064, or $64 for 1,000 such tasks.

That is 1.2 times cheaper than the median output price of frontier models ($14.5).

To estimate your own budget, use the agent cost calculator.

Context and capabilities

Its context window reaches 1,000,000 tokens, roughly 1,500 pages of text at 500 words per page: enough to load a whole book or a mid-sized codebase. A single answer can run to 64,000 tokens.

It accepts as input: text and image. Its knowledge stops at June 2025: beyond that, it needs web search or your documents.

It can call tools (functions, APIs), which makes it a possible foundation for an AI agent.

Who is it for?

Given its category, Gemini 3 Pro Preview is mainly aimed at hard tasks where quality matters more than cost: deep reasoning, complex code, long analyses.

Its Terminal-Bench result (69.4%, above the median of tracked models) makes it a serious candidate for coding agents.

Its very large context window makes it worth considering for analysing large sets of documents in a single request.

Before adopting it, test it on your own examples: benchmarks measure standard tasks, not your job.

Comparable models

In the same category, the models closest to Gemini 3 Pro Preview are:

• GPT-5.4 (OpenAI, index 87.2)
• GPT-5.6 Sol (OpenAI, index 87.3)
• Qwen3.7 Max (Alibaba Cloud (Qwen), index 86.6)
• Grok 4.6 (xAI, index 86.1)