MultimodalGoogle

Gemini 3 Pro

Made by AIWritten by AI (Claude) from the source below. Check anything important against the source.How it's made

Gemini 3 Pro — MMMU-Pro 81%, Video-MMMU 87.6%

Why it matters

A natively multimodal design paid off as an edge in video and document understanding

Source — Primary sources first; secondary reports are labelled “Press”

Officialblog.googlehttps://blog.google/products/gemini/gemini-3/

Added in Week 35, 2026 · See that week

On the map — ⑦ Multimodal line

  1. Previous stopDeepSeek-OCR
  2. This recordGemini 3 Pro
  3. Next stopKimi K2.5
See it on the map

In the model family tree — Gemini family · 21 versions

  1. Earlier versionGemini 2.5 Pro Experimental
  2. This recordGemini 3 Pro
  3. Later versionGemini 3.8 Flash / 3.8 Flash Cyber
See the whole Gemini family

Same week

Natural Emergent Misalignment from Reward HackingPaperAnthropic, OpenAI and MCP-UI propose the MCP Apps extension (SEP-1865)OfficialNano Banana Pro (Gemini 3 Pro Image)OfficialOlmo 3Official

Nearby on the same axis

Nano Banana Pro (Gemini 3 Pro Image)OfficialWorld Labs launches MarbleOfficialFLUX.2OfficialDeepSeek-OCRPaper